David Alvarez-Melis

(he/him/his)

Assistant Professor, Harvard University (SEAS)

150 Western Av. Room 2-332, Allston MA 02134

[three initials]@seas.harvard.edu

About

I'm an Assistant Professor of Computer Science at Harvard SEAS where I lead the Data-Centric Machine Learning (DCML) group. I'm also an Associate Faculty at the Kempner Institute, and have affiliations with the Harvard Data Science Initiative. I am also a researcher at Microsoft Research New England.

In my group, we study how data—not just models—shapes the behavior and reliability of AI systems. Our research develops foundational principles and methods for characterizing, transforming, and optimizing datasets to make learning more efficient, interpretable, and adaptive. By treating datasets as structured, dynamic entities rather than static inputs, we aim to build a rigorous science of data for machine learning: a framework that formally connects data properties to model outcomes. In other words, we seek to reivindicate the role of data in AI and ML, recognizing the reality that in this day and age data is often the most scarce, valuable, and manipulable resource in learning systems, yet we still lack a systematic understanding of how to harness it effectively, robustly, and predictably.

Our work integrates tools from optimal transport, information theory, and generative modeling, and applies them across domains including scientific data, language, and vision. For example, we have developed methods based on optimal transport to align datasets from different domains, compare complex data distributions, generate synthetic data that preserves key characteristics, and transform datasets to improve model robustness. We have also explored techniques for interpreting black-box models, enhancing their trustworthiness through robust explanations, and incorporating human-centered design principles.

If you share these goals, please reach out. We are always looking for collaborators and new group members!

Prospective lab members: If you are interested in joining my group at Harvard, please read this.

Bio

I obtained a PhD in computer science from MIT, where I worked at CSAIL on various topics in machine learning and natural language processing. I also hold BSc (Licenciatura) and MS degrees in mathematics from ITAM and Courant Institute (NYU), respectively. During the latter, I worked on semidefinite programming for domain adaptation for my thesis. Between Master's and PhD, I spent a year at IBM's T.J. Watson Research Center, working in the Speech Recognition and NLP teams.

News

[10/29/2026] Giving an invited talk at Harvard SEAS Computer Science Colloquium Series.
[10/5/2026] Giving an invited talk at Workshop on Probabilistic Modeling, Flatiron Institute & IBM Research.
[9/26/2026] Papers accepted to NeurIPS 2026: Bridging Compute- and Data-Optimal Pretraining (led by Sunny Qin); RL Excursions during Pre-Training: Re-examining Policy Optimization for LLM Training (led by Rachit Bansal, Clara Mohri & Sunny Qin); Understanding Layer Patching in Model Size Interpolation (led by Sara Kangaslahti);
[8/25/2026] Papers accepted to COLM 2026: Domain-Aware Scaling Laws Uncover Data Synergy (led by Kimia Hamidieh); Curriculum Learning as Transport: Understanding Curricula with Wasserstein Geodesics (led by Changho Shin);
[7/13/2026] Lab papers at three ICML 2026 workshops: AdaptFM, HiLD, and the Mechanistic Interpretability Workshop.
[6/3/2026] Giving an invited talk at SIAM Conference on Optimization (OP26), Edinburgh.
[4/30/2026] Papers accepted to ICML 2026: Random Scaling of Emergent Capabilities (led by Rosie Zhao); Progressive Unmasking Accelerates Masked Diffusion Training (led by Jay Kim & Jonathan Geuter); A Critical Look at Targeted Instruction Selection (led by Nihal Nayak);
[2/20/2026] Giving an invited talk at Carnegie Mellon University. Slides
[1/25/2026] Papers accepted to ICLR 2026: Boomerang Distillation (led by Sara Kangaslahti); Guided Speculative Inference (led by Jonathan Geuter); Test-Time Training (led by Rachit Bansal);
[1/22/2026] Paper: Three-operator splitting with stale gradients for faster non-linear optimal transport led by Jacob Lindbäck accepted to AISTATS 2026.
[12/06/2025] Giving an invited talk at UniReps Workshop @ NeurIPS 2025. Slides
[08/25/2025] Paper: Continuous Language Model Interpolation for Dynamic and Controllable Text Generation led by Sara Kangaslahti accepted to TMLR.
[08/20/2025] Paper: Investigating the interaction of linguistic and mathematical reasoning in language models using multilingual number puzzles led by Antara Bhattacharya accepted to EMNLP 2025.
[08/20/2025] Paper: Data Drives Unstable Hierarchical Generalization in LMs led by Sunny Qin accepted to EMNLP 2025.
[07/08/2025] Giving a keynote talk at AstroAI Workshop 2025.
[07/08/2025] Paper: To Backtrack or Not to Backtrack: When Sequential Search Limits Model Reasoning led by Sunny Qin accepted to COLM 2025.
[05/20/2025] Paper: What is the Right Notion of Distance between Predict-then-Optimize Tasks? led by Paula Rodriguez-Diaz accepted to UAI 2025.
[03/19/2025] OTDD has been incorporated into the DataSimilarity R package.
[01/22/2025] Paper: DDEQs: Distributional Deep Equilibrium Models through Wasserstein Gradient Flows led by Jonathan Geuter accepted to AISTATS 2025.
[01/22/2025] Paper: Mixture of Parrots: Experts improve memorization more than reasoning led by Samy Jelassi accepted to ICLR 2025.
[09/27/2024] Our son Noah was born!
[09/26/2024] Paper: A Label is Worth a Thousand Images in Dataset Distillation led by Sunny Qin accepted to NeurIPS 2024.
[07/26/2024] Giving a talk at ICML 2024 'Foundation Models in the Wild' workshop.
[07/23/2024] Attending ICML 2024 in Vienna. Our group is presenting work on LLMs, Optimal Transport, Model Interpolation and more!
[06/26/2024] Joined the Kempner Institute as Affiliate Faculty.
[05/01/2024] Paper: Tag-LLM: Repurposing General-Purpose LLMs for Specialized Domains led by MSR Summer Intern Junhong Shen accepted to ICML 2024.
[05/11/2023] Will be serving as AC for ACML 2023.
[05/08/2023] Paper: Generating Synthetic Datasets by Interpolating along Generalized Geodesics led by MSR Summer Intern Jiaojiao Fan accepted to UAI 2023.
[04/24/2023] Paper: InfoOT: Information Maximizing Optimal Transport led by MSR Summer Intern Ching-Yao Chuang accepted to ICML 2023.
[02/28/2022] Will be serving as AC for NeurIPS 2023.
[02/24/2023] Paper: Domain adaptation using optimal transport for invariant learning using histopathology datasets led by MSR Summer Intern Kia Falahkheirkhah accepted to MIDL 2023.
[10/06/2022] New Preprint: InfoOT: Information Maximizing Optimal Transport.
[09/30/2022] New Preprint: Neural Unbalanced Optimal Transport via Cycle-Consistent Semi-Couplings.
[09/15/2022] Paper: Are GANs overkill for NLP? accepted to NeurIPS 2022.
[07/15/2022] Paper: Optimizing Functionals on the Space of Probabilities with Input Convex Neural Networks accepted to TMLR.
[07/06/2022] Anna Yeaton will be preseting our work on Optimal Transport for Histopathology at MIDL 2022.
[05/16/2022] Giving a talk at the IMSI Workshop on Applied Optimal Transport: 'Machine Learning in the Space of Datasets: an Optimal Transport Perspective'
[05/14/2022] Will be serving as AC for ACML 2022.
[03/18/2022] Giving a talk at the AMS Special Session on Maths of Data Science: 'Principled Data Manipulation with Optimal Transport'
[01/31/2022] I will be serving as associate chair for ICML 2022.
[12/06/21-12/14/21] 'Attending' NeurIPS. Presenting 'JKO-ICNN' at the OTML workshop.
[11/15/2021] Guest lecture on NLP/Language Models at Harvard's CS182.
[10/28/2021] Talk: 'Principled Data Manipulation with OT', BIRS-CMO workshop on 'Geometry & Learning from Data'.
[08/21/2021] Paper: From Human Explanation to Model Interpretabilty: A Framework Based on Weight of Evidence accepted to HCOMP 2021.
[07/23/2021] Talk: 'Comparing, Transforming, and Optimizing Datasets with OT', ML4Data workshop at ICML.
[06/01/2021] New Preprint: Optimizing Functionals on the Space of Probabilities with Input Convex Neural Networks.
[05/08/2021] Paper: Dataset Dynamics via Gradient Flows in Probability Space accepted to ICML 2021.
[03/20/2021] Reviewer award at ICLR 2021.
[12/06/20-12/12/20] 'Attending' NeurIPS. Presenting 'Geometric Dataset Distances via Optimal Transport'.
[11/18/20] Giving a talk at MSR's Directions in ML series.
[09/28/20] Paper: Geometric Dataset Distances via Optimal Transport accepted to NeurIPS 2020.
[09/28/20] Optimal Transport Dataset Distance featured in the MSR blog.
[01/06/20] Paper: Unsupervised Hierarchy Matching with Optimal Transport over Hyperbolic Spaces accepted to AISTATS 2020.
[12/10-12/15] At NeurIPS 2019, presenting at the HCML and OTML Workshops
[10/31/19] New preprint on Interpretability via Weight of Evidence (to be presented at HCML@NeurIPS2019)
[09/01/19] Reviewer award (among 400 highest-scoring reviewers) at NeurIPS 2019.
[08/26/19] Started postdoc at MSR
[07/15/19] Final thesis submission - PhD completed!
[06/21/19] Succesfully defended my thesis!
[04/21/19] Paper: Functional Transparency for Structured Data: a Game-Theoretic Approach. accepted to ICML 2019.
[04/21/19] Paper: Learning Generative Models Across Incomparable Spaces accepted to ICML 2019.
[04/14-04/19] At AISTATS 2019.
[12/22/18] Paper: Towards Optimal Transport with Global Invariances accepted to AISTATS 2019.
[12/21/18] Paper: Towards Robust, Locally Linear Deep Networks accepted to ICLR 2019.
[12/08/18] Received best paper award in the R2L Workshop at NeurIPS. Congratulations Charlotte!
[12/02->09/18] At NeurIPS 2018.
[11/03/18] Talk on Gromov-Wasserstein embedding alignment @ EMNLP.
[10/30-11/04] At EMNLP 2018.
[10/30] Our EMNLP paper was covered by MIT NEWS.
[09/05/18] Paper: Towards Robust Interpretability with Self-Explaining Neural Networks accepted to NIPS 2018.
[09/03/18] Named one of 218 best reviewers at NIPS 2018. Thanks NIPS for the free registration!
[08/25-26/18] In Mexico City, for our student-run RIIAA conference.
[08/24/18] Finished Internship!
[08/10/18] Paper: Gromov-Wasserstein Alignment of Word Embedding Spaces accepted to EMNLP 2018.
[07/10-15/18] At ICML 2018.
[06/25/18] New Preprint: Towards Optimal Transport with Global Invariances.
[05/29/18] Started Internship at MSR NYC.
[04/17/18] Guest lecture on Interpretabilty for NLP @ CMU ECE739.
[04/11/18] Presenting our work on Structured OT @ AISTATS.
[02/01/18] Named finalist for Facebook PhD Fellowship.
[01/12/18] Talk at OpenAI on interpretability + optimal transport.
[11/16/18] Talk at CompLang Seminar @ MIT on interpretability in NLP.

Publications

Most recent publications on Google Scholar.

Bridging Compute- and Data-Optimal Pretraining

Tian Qin, Kimia Hamidieh, David Alvarez-Melis

NeurIPS'26: Neural Information Processing Systems. 2026. To appear. Oral.

pdf arxiv code bib

 @inproceedings{qin2026bridging,
  title         = "Bridging Compute- and Data-Optimal Pretraining",
  author        = "Qin, Tian and Hamidieh, Kimia and Alvarez-Melis, David",
  booktitle     = "Advances in Neural Information Processing Systems",
  year          = "2026",
  eprint        = "2607.25271",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2607.25271"
}

Let's (not) just put things in Context: Test-time Training for Long-context LLMs

Rachit Bansal, Aston Zhang, Rishabh Tiwari, Lovish Madaan, Venkata Sai Surya Subramanyam Duvvuri, Devvrit Khatri, David Brandfonbrener, David Alvarez-Melis, Prajjwal Bhargava, Mihir Kale, Samy Jelassi

ICLR'26: International Conference on Learning Representations. 2026.

arxiv openreview bib

 @inproceedings{bansal2026lets,
  title         = "Let's (not) just put things in Context: Test-time Training for Long-context {LLMs}",
  author        = "Bansal, Rachit and Zhang, Aston and Tiwari, Rishabh and Madaan, Lovish and Duvvuri, Venkata Sai Surya Subramanyam and Khatri, Devvrit and Brandfonbrener, David and Alvarez-Melis, David and Bhargava, Prajjwal and Kale, Mihir and Jelassi, Samy",
  booktitle     = "International Conference on Learning Representations",
  year          = "2026",
  eprint        = "2512.13898",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2512.13898"
}

RL Excursions during Pre-training: How early is too early for on-policy learning?

Rachit Bansal, Clara Mohri, Tian Qin, David Alvarez-Melis, Sham Kakade

NeurIPS'26: Neural Information Processing Systems. 2026. To appear.

arxiv project bib

 @inproceedings{bansal2026rl,
  title         = "{RL} Excursions during Pre-training: How early is too early for on-policy learning?",
  author        = "Bansal, Rachit and Mohri, Clara and Qin, Tian and Alvarez-Melis, David and Kakade, Sham",
  booktitle     = "Advances in Neural Information Processing Systems",
  year          = "2026",
  eprint        = "2606.04272",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2606.04272"
}

Guided Speculative Inference for Efficient Test-Time Alignment of LLMs

Jonathan Geuter, Youssef Mroueh, David Alvarez-Melis

ICLR'26: International Conference on Learning Representations. 2026.

arxiv openreview code bib

 @inproceedings{geuter2025guided,
  title         = "Guided Speculative Inference for Efficient Test-Time Alignment of {LLMs}",
  author        = "Geuter, Jonathan and Mroueh, Youssef and Alvarez-Melis, David",
  booktitle     = "International Conference on Learning Representations",
  year          = "2026",
  eprint        = "2506.04118",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2506.04118"
}

Domain-Aware Scaling Laws Uncover Data Synergy

Kimia Hamidieh, Lester Mackey, David Alvarez-Melis

COLM'26: Conference on Language Modeling. 2026.

arxiv openreview bib

 @inproceedings{hamidieh2026domain-aware,
  title         = "Domain-Aware Scaling Laws Uncover Data Synergy",
  author        = "Hamidieh, Kimia and Mackey, Lester and Alvarez-Melis, David",
  booktitle     = "Conference on Language Modeling",
  year          = "2026",
  eprint        = "2607.11052",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2607.11052"
}

Boomerang Distillation Enables Zero-Shot Model Size Interpolation

Sara Kangaslahti, Nihal V. Nayak, Jonathan Geuter, Marco Fumero, Francesco Locatello, David Alvarez-Melis

ICLR'26: International Conference on Learning Representations. 2026.

arxiv openreview poster blog code bib

 @inproceedings{kangaslahti2026boomerang,
  title         = "Boomerang Distillation Enables Zero-Shot Model Size Interpolation",
  author        = "Kangaslahti, Sara and Nayak, Nihal V. and Geuter, Jonathan and Fumero, Marco and Locatello, Francesco and Alvarez-Melis, David",
  booktitle     = "International Conference on Learning Representations",
  year          = "2026",
  eprint        = "2510.05064",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2510.05064"
}

Understanding Layer Patching in Model Size Interpolation

Sara Kangaslahti, Jonathan Geuter, Nihal V. Nayak, Marco Fumero, Francesco Locatello, David Alvarez-Melis

NeurIPS'26: Neural Information Processing Systems. 2026. To appear.

arxiv bib

 @inproceedings{kangaslahti2026layer,
  title         = "Understanding Layer Patching in Model Size Interpolation",
  author        = "Kangaslahti, Sara and Geuter, Jonathan and Nayak, Nihal V. and Fumero, Marco and Locatello, Francesco and Alvarez-Melis, David",
  booktitle     = "Advances in Neural Information Processing Systems",
  year          = "2026",
  eprint        = "2607.08170",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2607.08170"
}

Stop Training for the Worst: Progressive Unmasking Accelerates Masked Diffusion Training

Jaeyeon Kim, Jonathan Geuter, David Alvarez-Melis, Sham M. Kakade, Sitan Chen

ICML'26: International Conference on Machine Learning. 2026.

arxiv openreview code bib

 @inproceedings{kim2026stop,
  title         = "Stop Training for the Worst: Progressive Unmasking Accelerates Masked Diffusion Training",
  author        = "Kim, Jaeyeon and Geuter, Jonathan and Alvarez-Melis, David and Kakade, Sham M. and Chen, Sitan",
  booktitle     = "Proceedings of the 43rd International Conference on Machine Learning",
  volume        = "306",
  pages         = "57515--57538",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2026",
  eprint        = "2602.10314",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2602.10314"
}

Three-operator Splitting with Stale Gradients for Faster Non-linear Optimal Transport

Jacob Lindbäck, David Alvarez-Melis, Mikael Johansson

AISTATS'26: International Conference on Artificial Intelligence and Statistics. 2026.

openreview poster bib

 @inproceedings{lindback2026three-operator,
  title     = "Three-operator Splitting with Stale Gradients for Faster Non-linear Optimal Transport",
  author    = "Lindbäck, Jacob and Alvarez-Melis, David and Johansson, Mikael",
  booktitle = "Proceedings of The 29th International Conference on Artificial Intelligence and Statistics",
  volume    = "300",
  publisher = "PMLR",
  series    = "Proceedings of Machine Learning Research",
  year      = "2026"
}

A Critical Look at Targeted Instruction Selection: Disentangling What Matters (and What Doesn't)

Nihal V. Nayak, Paula Rodriguez-Diaz, Neha Hulkund, Sara Beery, David Alvarez-Melis

ICML'26: International Conference on Machine Learning. 2026.

arxiv openreview code bib

 @inproceedings{nayak2026critical,
  title         = "A Critical Look at Targeted Instruction Selection: Disentangling What Matters (and What Doesn't)",
  author        = "Nayak, Nihal V. and Rodriguez-Diaz, Paula and Hulkund, Neha and Beery, Sara and Alvarez-Melis, David",
  booktitle     = "Proceedings of the 43rd International Conference on Machine Learning",
  volume        = "306",
  pages         = "92008--92043",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2026",
  eprint        = "2602.14696",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2602.14696"
}

Curriculum Learning as Transport: Understanding Curricula with Wasserstein Geodesics

Changho Shin, David Alvarez-Melis

COLM'26: Conference on Language Modeling. 2026.

arxiv openreview bib

 @inproceedings{shin2026curriculum,
  title         = "Curriculum Learning as Transport: Understanding Curricula with Wasserstein Geodesics",
  author        = "Shin, Changho and Alvarez-Melis, David",
  booktitle     = "Conference on Language Modeling",
  year          = "2026",
  eprint        = "2609.09099",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2609.09099"
}

Integrating chemical structures as treatments improves representations of microscopy images for morphological profiling

Yemin Yu, Emre Hayir, Neil Tenenholtz, Lester Mackey, Ying Wei, David Alvarez-Melis, Ava P. Amini, Alex X. Lu

PLOS Computational Biology. 2026.

arxiv code dataset bib

 @article{yu2025causal,
  title         = "Integrating chemical structures as treatments improves representations of microscopy images for morphological profiling",
  author        = "Yu, Yemin and Hayir, Emre and Tenenholtz, Neil and Mackey, Lester and Wei, Ying and Alvarez-Melis, David and Amini, Ava P. and Lu, Alex X.",
  journal       = "PLOS Computational Biology",
  volume        = "22",
  number        = "7",
  doi           = "10.1371/journal.pcbi.1014483",
  year          = "2026",
  eprint        = "2504.09544",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2504.09544"
}

Random Scaling of Emergent Capabilities

Rosie Zhao, Tian Qin, David Alvarez-Melis, Sham M. Kakade, Naomi Saphra

ICML'26: International Conference on Machine Learning. 2026.

arxiv openreview slides bib

 @inproceedings{zhao2025distributional,
  title         = "Random Scaling of Emergent Capabilities",
  author        = "Zhao, Rosie and Qin, Tian and Alvarez-Melis, David and Kakade, Sham M. and Saphra, Naomi",
  booktitle     = "Proceedings of the 43rd International Conference on Machine Learning",
  volume        = "306",
  pages         = "161980--161999",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2026",
  eprint        = "2502.17356",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2502.17356"
}

Thinking at the Right Size: Amortized Distillation Across Post-Trained LLMs

Yan Zhou, Sara Kangaslahti, Jonathan Geuter, Nihal V. Nayak, Marco Fumero, Francesco Locatello, David Alvarez-Melis

EMNLP Findings'26: Findings of the Association for Computational Linguistics: EMNLP. 2026.

arxiv code bib

 @inproceedings{zhou2026thinking,
  title         = "Thinking at the Right Size: Amortized Distillation Across Post-Trained {LLMs}",
  author        = "Zhou, Yan and Kangaslahti, Sara and Geuter, Jonathan and Nayak, Nihal V. and Fumero, Marco and Locatello, Francesco and Alvarez-Melis, David",
  booktitle     = "Findings of the Association for Computational Linguistics: EMNLP 2026",
  year          = "2026",
  eprint        = "2608.22854",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2608.22854"
}

Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention

Jing Huang, Daniel Wurgaft, Rachit Bansal, Laura Ruis, Naomi Saphra, David Alvarez-Melis, Andrew Kyle Lampinen, Christopher Potts, Ekdeep Singh Lubana

ICML 2026 Workshop on High Dimensional Learning Dynamics.

arxiv bib

 @inproceedings{huang2026why-hild,
  title         = "Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention",
  author        = "Huang, Jing and Wurgaft, Daniel and Bansal, Rachit and Ruis, Laura and Saphra, Naomi and Alvarez-Melis, David and Lampinen, Andrew Kyle and Potts, Christopher and Lubana, Ekdeep Singh",
  booktitle     = "ICML 2026 Workshop on High Dimensional Learning Dynamics",
  year          = "2026",
  eprint        = "2605.29548",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2605.29548"
}

Can Interpretation Predict Behavior on Unseen Data?

Victoria R. Li, Jenny Kaufmann, Tian Qin, Martin Wattenberg, David Alvarez-Melis, Naomi Saphra

Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP, 2026.

arxiv code bib

 @inproceedings{li2025interpretation,
  title         = "Can Interpretation Predict Behavior on Unseen Data?",
  author        = "Li, Victoria R. and Kaufmann, Jenny and Qin, Tian and Wattenberg, Martin and Alvarez-Melis, David and Saphra, Naomi",
  booktitle     = "Proceedings of the 9th {BlackboxNLP} Workshop: Analyzing and Interpreting Neural Networks for {NLP}",
  year          = "2026",
  eprint        = "2507.06445",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2507.06445"
}

Adam or Gauss-Newton? A Comparative Study In Terms of Basis Alignment and SGD Noise

Bingbin Liu, Rachit Bansal, Depen Morwani, David Alvarez-Melis, Sham M. Kakade

ICML 2026 Workshop on High Dimensional Learning Dynamics.

arxiv openreview bib

 @inproceedings{liu2026adam-hild,
  title         = "Adam or Gauss-Newton? A Comparative Study In Terms of Basis Alignment and SGD Noise",
  author        = "Liu, Bingbin and Bansal, Rachit and Morwani, Depen and Alvarez-Melis, David and Kakade, Sham M.",
  booktitle     = "ICML 2026 Workshop on High Dimensional Learning Dynamics",
  year          = "2026",
  eprint        = "2510.13680",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2510.13680"
}

Adam or Gauss-Newton? A Comparative Study In Terms of Basis Alignment and SGD Noise

Bingbin Liu, Rachit Bansal, Depen Morwani, Nikhil Vyas, David Alvarez-Melis, Sham M. Kakade

NeurIPS 2026 Workshop on Optimization for Machine Learning. To appear.

arxiv openreview bib

 @inproceedings{liu2026adam-opt,
  title         = "Adam or Gauss-Newton? A Comparative Study In Terms of Basis Alignment and SGD Noise",
  author        = "Liu, Bingbin and Bansal, Rachit and Morwani, Depen and Vyas, Nikhil and Alvarez-Melis, David and Kakade, Sham M.",
  booktitle     = "NeurIPS 2026 Workshop on Optimization for Machine Learning",
  year          = "2026",
  eprint        = "2510.13680",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2510.13680"
}

Lossy Superposition: Predicting Compositional Errors Without Seeing Inputs

Jennifer Meng Lu, Ruochen Zhang, Isabelle Lee, David Alvarez-Melis, Ellie Pavlick, Naomi Saphra

ICML 2026 Workshop on Mechanistic Interpretability.

openreview bib

 @inproceedings{lu2026lossy,
  title     = "Lossy Superposition: Predicting Compositional Errors Without Seeing Inputs",
  author    = "Lu, Jennifer Meng and Zhang, Ruochen and Lee, Isabelle and Alvarez-Melis, David and Pavlick, Ellie and Saphra, Naomi",
  booktitle = "ICML 2026 Workshop on Mechanistic Interpretability",
  year      = "2026"
}

Feeding BabyLMs Macaroni: Code-Switching Curricula Cause Cross-Lingual Convergence

Dries Rooryck, Alexander Dazhen Cai, Yonatan Belinkov, David Alvarez-Melis, Kianté Brantley

Proceedings of the Second BabyLM Workshop, 2026.

arxiv code bib

 @inproceedings{rooryck2026feeding,
  title         = "Feeding {BabyLMs} Macaroni: Code-Switching Curricula Cause Cross-Lingual Convergence",
  author        = "Rooryck, Dries and Cai, Alexander Dazhen and Belinkov, Yonatan and Alvarez-Melis, David and Brantley, Kianté",
  booktitle     = "Proceedings of the Second {BabyLM} Workshop",
  year          = "2026",
  eprint        = "2609.30535",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2609.30535"
}

Beyond Explainable AI (XAI): An Overdue Paradigm Shift and Post-XAI Research Directions

Saleh Afroogh, Syed Ishtiaque Ahmed, Petra Ahrweiler, David Alvarez-Melis, Mansur Maturidi Arief, Emilia Barakova, Falco J. Bargagli-Stoffi, Erdem Biyik, Hanjie Chen, Xiang 'Anthony' Chen, Robert Alan Clements, Keeley Crockett, Amit Dhurandhar, Fethiye Irmak Dogan, Mollie Dollinger, Motahhare Eslami, Aldo A Faisal, Arya Farahi, Melanie F. Pradier, Saadia Gabriel, Diego Garcia-Olano, Marzyeh Ghassemi, Shaona Ghosh, Hatice Gunes, Ehsan Hajiramezanali, Stefan Haufe, Biwei Huang, Angel Hwang, Md Tauhidul Islam, Junfeng Jiao, Amir-Hossein Karimi, Saber Kazeminasab, Anastasia Kuzminykh, William La Cava, Brian Y. Lim, Xiaofeng Liu, Mohammad R. K. Mofrad, Alicia Parrish, Maria Perez-Ortiz, Shriti Raj, Swabha Swayamdipta, Salmonn Talebi, Kush R. Varshney, Mihaela Vorvoreanu, Lily Weng, Alice Xiang, Yiming Xu, Ding Zhao, Jieyu Zhao

arXiv preprint. 2026.

arxiv bib

 @misc{afroogh2026beyond,
  title         = "Beyond Explainable AI (XAI): An Overdue Paradigm Shift and Post-XAI Research Directions",
  author        = "Afroogh, Saleh and Ahmed, Syed Ishtiaque and Ahrweiler, Petra and Alvarez-Melis, David and Arief, Mansur Maturidi and Barakova, Emilia and Bargagli-Stoffi, Falco J. and Biyik, Erdem and Chen, Hanjie and Chen, Xiang 'Anthony' and Clements, Robert Alan and Crockett, Keeley and Dhurandhar, Amit and Dogan, Fethiye Irmak and Dollinger, Mollie and Eslami, Motahhare and Faisal, Aldo A and Farahi, Arya and Pradier, Melanie F. and Gabriel, Saadia and Garcia-Olano, Diego and Ghassemi, Marzyeh and Ghosh, Shaona and Gunes, Hatice and Hajiramezanali, Ehsan and Haufe, Stefan and Huang, Biwei and Hwang, Angel and Islam, Md Tauhidul and Jiao, Junfeng and Karimi, Amir-Hossein and Kazeminasab, Saber and Kuzminykh, Anastasia and La Cava, William and Lim, Brian Y. and Liu, Xiaofeng and Mofrad, Mohammad R. K. and Parrish, Alicia and Perez-Ortiz, Maria and Raj, Shriti and Swayamdipta, Swabha and Talebi, Salmonn and Varshney, Kush R. and Vorvoreanu, Mihaela and Weng, Lily and Xiang, Alice and Xu, Yiming and Zhao, Ding and Zhao, Jieyu",
  year          = "2026",
  eprint        = "2602.24176",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2602.24176"
}

Releasing the Inner Jev: A Simple Recipe for Open Decision Models

Jonathan Geuter, Andrew Cheng, David Alvarez-Melis

arXiv preprint. 2026.

arxiv code dataset bib

 @misc{geuter2026releasing,
  title         = "Releasing the Inner Jev: A Simple Recipe for Open Decision Models",
  author        = "Geuter, Jonathan and Cheng, Andrew and Alvarez-Melis, David",
  year          = "2026",
  eprint        = "2610.17205",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2610.17205"
}

Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention

Jing Huang, Daniel Wurgaft, Rachit Bansal, Laura Ruis, Naomi Saphra, David Alvarez-Melis, Andrew Kyle Lampinen, Christopher Potts, Ekdeep Singh Lubana

arXiv preprint. 2026.

arxiv bib

 @misc{huang2026why,
  title         = "Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention",
  author        = "Huang, Jing and Wurgaft, Daniel and Bansal, Rachit and Ruis, Laura and Saphra, Naomi and Alvarez-Melis, David and Lampinen, Andrew Kyle and Potts, Christopher and Lubana, Ekdeep Singh",
  year          = "2026",
  eprint        = "2605.29548",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2605.29548"
}

Adversarial Concept Search: Predicting Compositional Errors From Feature Geometry

Jennifer Meng Lu, Ruochen Zhang, Isabelle Lee, David Alvarez-Melis, Ellie Pavlick, Naomi Saphra

arXiv preprint. 2026.

arxiv bib

 @misc{lu2026adversarial,
  title         = "Adversarial Concept Search: Predicting Compositional Errors From Feature Geometry",
  author        = "Lu, Jennifer Meng and Zhang, Ruochen and Lee, Isabelle and Alvarez-Melis, David and Pavlick, Ellie and Saphra, Naomi",
  year          = "2026",
  eprint        = "2606.13934",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2606.13934"
}

Universal Bounds for Self-Normalized Kullback-Leibler Estimation

Buu Phan, Lester Mackey, David Alvarez-Melis

arXiv preprint. 2026.

arxiv bib

 @misc{phan2026universal,
  title         = "Universal Bounds for Self-Normalized Kullback-Leibler Estimation",
  author        = "Phan, Buu and Mackey, Lester and Alvarez-Melis, David",
  year          = "2026",
  eprint        = "2610.12851",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2610.12851"
}

Low-Frequency Shortcuts in Texture-Driven Visual Learning

Utku Şirin, Cathy Hou, Despina-Ekaterini Argiropoulos, David Alvarez-Melis, Stratos Idreos

arXiv preprint. 2026.

arxiv bib

 @misc{sirin2026low-frequency,
  title         = "Low-Frequency Shortcuts in Texture-Driven Visual Learning",
  author        = "Şirin, Utku and Hou, Cathy and Argiropoulos, Despina-Ekaterini and Alvarez-Melis, David and Idreos, Stratos",
  year          = "2026",
  eprint        = "2606.03493",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2606.03493"
}

OT on the Map: Quantifying Domain Shifts in Geographic Space

Haoran Zhang, Livia Betti, Konstantin Klemmer, Esther Rolf, David Alvarez-Melis

arXiv preprint. 2026.

arxiv code bib

 @misc{zhang2026ot,
  title         = "OT on the Map: Quantifying Domain Shifts in Geographic Space",
  author        = "Zhang, Haoran and Betti, Livia and Klemmer, Konstantin and Rolf, Esther and Alvarez-Melis, David",
  year          = "2026",
  eprint        = "2604.16220",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2604.16220"
}

Investigating the interaction of linguistic and mathematical reasoning in language models using multilingual number puzzles

Antara Raaghavi Bhattacharya, Isabel Papadimitriou, Kathryn Davidson, David Alvarez-Melis

EMNLP'25: Empirical Methods in Natural Language Processing. 2025.

arxiv poster talk code bib

 @inproceedings{bhattacharya2025investigating,
  title         = "Investigating the interaction of linguistic and mathematical reasoning in language models using multilingual number puzzles",
  author        = "Bhattacharya, Antara Raaghavi and Papadimitriou, Isabel and Davidson, Kathryn and Alvarez-Melis, David",
  booktitle     = "Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing",
  pages         = "28322--28332",
  year          = "2025",
  eprint        = "2506.13886",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2506.13886"
}

scGeneScope: A Treatment-Matched Single Cell Imaging and Transcriptomics Dataset and Benchmark for Treatment Response Modeling

Joel Dapello, Marcel Nassar, Ridvan Eksi, Ban Wang, Jules Gagnon-Marchand, Kenneth Gao, Akram Baharlouei, Kyra Thrush, Nina Riehs, Amy Peterson, Aniket Tolpadi, Abhejit Rajagopal, Henry Miller, Ashley Mae Conard, David Alvarez-Melis, Rory Stark, Simone Bianco, Morgan Levine, Ava P. Amini, Alex X. Lu, Nicolò Fusi, Ravi Pandya, Valentina Pedoia, Hana El-Samad

NeurIPS D&B'25: Neural Information Processing Systems, Datasets and Benchmarks Track. 2025.

openreview code bib

 @inproceedings{dapello2025scgenescope,
  title     = "{scGeneScope}: A Treatment-Matched Single Cell Imaging and Transcriptomics Dataset and Benchmark for Treatment Response Modeling",
  author    = "Dapello, Joel and Nassar, Marcel and Eksi, Ridvan and Wang, Ban and Gagnon-Marchand, Jules and Gao, Kenneth and Baharlouei, Akram and Thrush, Kyra and Riehs, Nina and Peterson, Amy and Tolpadi, Aniket and Rajagopal, Abhejit and Miller, Henry and Conard, Ashley Mae and Alvarez-Melis, David and Stark, Rory and Bianco, Simone and Levine, Morgan and Amini, Ava P. and Lu, Alex X. and Fusi, Nicolò and Pandya, Ravi and Pedoia, Valentina and El-Samad, Hana",
  booktitle = "Advances in Neural Information Processing Systems",
  volume    = "38",
  year      = "2025"
}

DDEQs: Distributional Deep Equilibrium Models through Wasserstein Gradient Flows

Jonathan Geuter, Clément Bonet, Anna Korba, David Alvarez-Melis

AISTATS'25: International Conference on Artificial Intelligence and Statistics. 2025.

arxiv openreview code bib

 @inproceedings{geuter2025distributional,
  title         = "{DDEQs}: Distributional Deep Equilibrium Models through {Wasserstein} Gradient Flows",
  author        = "Geuter, Jonathan and Bonet, Clément and Korba, Anna and Alvarez-Melis, David",
  booktitle     = "Proceedings of The 28th International Conference on Artificial Intelligence and Statistics",
  volume        = "258",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2025",
  eprint        = "2503.01140",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2503.01140"
}

Mixture of Parrots: Experts improve memorization more than reasoning

Samy Jelassi, Clara Mohri, David Brandfonbrener, Alex Gu, Nikhil Vyas, Nikhil Anand, David Alvarez-Melis, Yuanzhi Li, Sham M. Kakade, Eran Malach

ICLR'25: International Conference on Learning Representations. 2025.

arxiv openreview slides blog bib

 @inproceedings{jelassi2025mixture,
  title         = "Mixture of Parrots: Experts improve memorization more than reasoning",
  author        = "Jelassi, Samy and Mohri, Clara and Brandfonbrener, David and Gu, Alex and Vyas, Nikhil and Anand, Nikhil and Alvarez-Melis, David and Li, Yuanzhi and Kakade, Sham M. and Malach, Eran",
  booktitle     = "International Conference on Learning Representations",
  pages         = "37030--37056",
  year          = "2025",
  eprint        = "2410.19034",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2410.19034"
}

Continuous Language Model Interpolation yields Dynamic and Controllable Text Generation

Sara Kangaslahti, David Alvarez-Melis

TMLR: Transactions on Machine Learning Research. 2025.

arxiv openreview code bib

 @article{kangaslahti2024continuous,
  title         = "Continuous Language Model Interpolation yields Dynamic and Controllable Text Generation",
  author        = "Kangaslahti, Sara and Alvarez-Melis, David",
  journal       = "Transactions on Machine Learning Research",
  year          = "2025",
  eprint        = "2404.07117",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2404.07117"
}

JONES-19: A Cultural Image Dataset Based on The Grammar of Ornament

Linh Pham, Stuart M. Shieber, David Alvarez-Melis, Alexandros Haridis

KUI'25: International Conference on Culture and Computer Science. 2025.

doi dataset bib

 @inproceedings{pham2025jones,
  title     = "{JONES-19}: A Cultural Image Dataset Based on {The Grammar of Ornament}",
  author    = "Pham, Linh and Shieber, Stuart M. and Alvarez-Melis, David and Haridis, Alexandros",
  booktitle = "Proceedings of the 22nd International Conference on Culture and Computer Science: Remixing Analog and Digital",
  year      = "2025"
}

To Backtrack or Not to Backtrack: When Sequential Search Limits Model Reasoning

Tian Qin, David Alvarez-Melis, Samy Jelassi, Eran Malach

COLM'25: Conference on Language Modeling. 2025.

arxiv blog code bib

 @inproceedings{qin2025backtrack,
  title         = "To Backtrack or Not to Backtrack: When Sequential Search Limits Model Reasoning",
  author        = "Qin, Tian and Alvarez-Melis, David and Jelassi, Samy and Malach, Eran",
  booktitle     = "Conference on Language Modeling",
  year          = "2025",
  eprint        = "2504.07052",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2504.07052"
}

Data Drives Unstable Hierarchical Generalization in LMs

Tian Qin, Naomi Saphra, David Alvarez-Melis

EMNLP'25: Empirical Methods in Natural Language Processing. 2025.

arxiv code bib

 @inproceedings{qin2025data,
  title         = "Data Drives Unstable Hierarchical Generalization in {LMs}",
  author        = "Qin, Tian and Saphra, Naomi and Alvarez-Melis, David",
  booktitle     = "Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing",
  pages         = "11722--11740",
  year          = "2025",
  eprint        = "2412.04619",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2412.04619"
}

What is the Right Notion of Distance between Predict-then-Optimize Tasks?

Paula Rodriguez-Diaz, Lingkai Kong, Kai Wang, David Alvarez-Melis, Milind Tambe

UAI'25: Uncertainty in Artificial Intelligence. 2025.

pdf arxiv poster code bib

 @inproceedings{rodriguez-diaz2025what,
  title         = "What is the Right Notion of Distance between Predict-then-Optimize Tasks?",
  author        = "Rodriguez-Diaz, Paula and Kong, Lingkai and Wang, Kai and Alvarez-Melis, David and Tambe, Milind",
  booktitle     = "Proceedings of the Forty-first Conference on Uncertainty in Artificial Intelligence",
  volume        = "286",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2025",
  eprint        = "2409.06997",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2409.06997"
}

Reliable and Responsible Foundation Models: A Comprehensive Survey

Xinyu Yang, Junlin Han, Rishi Bommasani, Jinqi Luo, Wenjie Qu, Wangchunshu Zhou, Adel Bibi, Xiyao Wang, Jaehong Yoon, Elias Stengel-Eskin, Shengbang Tong, Lingfeng Shen, Rafael Rafailov, Runjia Li, Zhaoyang Wang, Yiyang Zhou, Chenhang Cui, Yu Wang, Wenhao Zheng, Huichi Zhou, Jindong Gu, Zhaorun Chen, Peng Xia, Tony Lee, Thomas P. Zollo, Vikash Sehwag, Jixuan Leng, Jiuhai Chen, Yuxin Wen, Huan Zhang, Zhun Deng, Linjun Zhang, Pavel Izmailov, Pang Wei Koh, Yulia Tsvetkov, Andrew Gordon Wilson, Jiaheng Zhang, James Zou, Cihang Xie, Hao Wang, Philip Torr, Julian McAuley, David Alvarez-Melis, Florian Tramèr, Kaidi Xu, Suman Jana, Chris Callison-Burch, René Vidal, Filippos Kokkinos, Mohit Bansal, Beidi Chen, Huaxiu Yao

TMLR: Transactions on Machine Learning Research. 2025.

arxiv openreview bib

 @article{yang2025reliable,
  title         = "Reliable and Responsible Foundation Models: A Comprehensive Survey",
  author        = "Yang, Xinyu and Han, Junlin and Bommasani, Rishi and Luo, Jinqi and Qu, Wenjie and Zhou, Wangchunshu and Bibi, Adel and Wang, Xiyao and Yoon, Jaehong and Stengel-Eskin, Elias and Tong, Shengbang and Shen, Lingfeng and Rafailov, Rafael and Li, Runjia and Wang, Zhaoyang and Zhou, Yiyang and Cui, Chenhang and Wang, Yu and Zheng, Wenhao and Zhou, Huichi and Gu, Jindong and Chen, Zhaorun and Xia, Peng and Lee, Tony and Zollo, Thomas P. and Sehwag, Vikash and Leng, Jixuan and Chen, Jiuhai and Wen, Yuxin and Zhang, Huan and Deng, Zhun and Zhang, Linjun and Izmailov, Pavel and Koh, Pang Wei and Tsvetkov, Yulia and Wilson, Andrew Gordon and Zhang, Jiaheng and Zou, James and Xie, Cihang and Wang, Hao and Torr, Philip and McAuley, Julian and Alvarez-Melis, David and Tramèr, Florian and Xu, Kaidi and Jana, Suman and Callison-Burch, Chris and Vidal, René and Kokkinos, Filippos and Bansal, Mohit and Chen, Beidi and Yao, Huaxiu",
  journal       = "Transactions on Machine Learning Research",
  year          = "2025",
  eprint        = "2602.08145",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2602.08145"
}

IGDA: Interactive Graph Discovery through Large Language Model Agents

Alexander Havrilla, David Alvarez-Melis, Nicolò Fusi

ICLR 2025 Workshop on Reasoning and Planning for Large Language Models.

arxiv openreview code bib

 @inproceedings{havrilla2025igda,
  title         = "{IGDA}: Interactive Graph Discovery through Large Language Model Agents",
  author        = "Havrilla, Alexander and Alvarez-Melis, David and Fusi, Nicolò",
  booktitle     = "ICLR 2025 Workshop on Reasoning and Planning for Large Language Models",
  year          = "2025",
  eprint        = "2502.17189",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2502.17189"
}

Bispectral OT: Dataset Comparison using Symmetry-Aware Optimal Transport

Annabel Ma, Kaiying Hou, David Alvarez-Melis, Melanie Weber

NeurIPS 2025 Workshop on Symmetry and Geometry in Neural Representations.

arxiv openreview code bib

 @inproceedings{ma2025bispectral,
  title         = "Bispectral OT: Dataset Comparison using Symmetry-Aware Optimal Transport",
  author        = "Ma, Annabel and Hou, Kaiying and Alvarez-Melis, David and Weber, Melanie",
  booktitle     = "NeurIPS 2025 Workshop on Symmetry and Geometry in Neural Representations",
  year          = "2025",
  eprint        = "2509.20678",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2509.20678"
}

Augmenting X-ray Astronomical Representations with Scientific Knowledge through Contrastive Learning

Juan Rafael Martínez-Galarza, Nicolò Oreste Pinciroli Vago, Shivam Raval, Carolina Cuesta-Lazaro, Melanie Weber, David Alvarez-Melis, Alberto Accomazzi, Cecilia Garraffo, Joshua Knutson, Ryan Thill, Christopher B. Green, Imantha Ahangama

ICLR 2025 Second Workshop on Representational Alignment.

openreview bib

 @inproceedings{martinez-galarza2025augmenting,
  title     = "Augmenting {X}-ray Astronomical Representations with Scientific Knowledge through Contrastive Learning",
  author    = "Martínez-Galarza, Juan Rafael and Vago, Nicolò Oreste Pinciroli and Raval, Shivam and Cuesta-Lazaro, Carolina and Weber, Melanie and Alvarez-Melis, David and Accomazzi, Alberto and Garraffo, Cecilia and Knutson, Joshua and Thill, Ryan and Green, Christopher B. and Ahangama, Imantha",
  booktitle = "ICLR 2025 Second Workshop on Representational Alignment",
  year      = "2025"
}

Adam or Gauss-Newton? A Comparative Study In Terms of Basis Alignment and SGD Noise

Bingbin Liu, Rachit Bansal, Depen Morwani, Nikhil Vyas, David Alvarez-Melis, Sham M. Kakade

arXiv preprint. 2025.

arxiv bib

 @misc{liu2025adam,
  title         = "Adam or Gauss-Newton? A Comparative Study In Terms of Basis Alignment and SGD Noise",
  author        = "Liu, Bingbin and Bansal, Rachit and Morwani, Depen and Vyas, Nikhil and Alvarez-Melis, David and Kakade, Sham M.",
  year          = "2025",
  eprint        = "2510.13680",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2510.13680"
}

Decomposing Elements of Problem Solving: What “Math” Does RL Teach?

Tian Qin, Core Francisco Park, Mujin Kwun, Aaron Walsman, Eran Malach, Nikhil Anand, Hidenori Tanaka, David Alvarez-Melis

arXiv preprint. 2025.

arxiv code bib

 @misc{qin2025decomposing,
  title         = "Decomposing Elements of Problem Solving: What ``Math'' Does {RL} Teach?",
  author        = "Qin, Tian and Park, Core Francisco and Kwun, Mujin and Walsman, Aaron and Malach, Eran and Anand, Nikhil and Tanaka, Hidenori and Alvarez-Melis, David",
  year          = "2025",
  eprint        = "2505.22756",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2505.22756"
}

KerJEPA: Kernel Discrepancies for Euclidean Self-Supervised Learning

Eric Zimmermann, Harley Wiltzer, Justin Szeto, David Alvarez-Melis, Lester Mackey

arXiv preprint. 2025.

arxiv bib

 @misc{zimmermann2025kerjepa,
  title         = "KerJEPA: Kernel Discrepancies for Euclidean Self-Supervised Learning",
  author        = "Zimmermann, Eric and Wiltzer, Harley and Szeto, Justin and Alvarez-Melis, David and Mackey, Lester",
  year          = "2025",
  eprint        = "2512.19605",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2512.19605"
}

A Label is Worth A Thousand Images in Dataset Distillation

Tian Qin, Zhiwei Deng, David Alvarez-Melis

NeurIPS'24: Neural Information Processing Systems. 2024.

pdf arxiv openreview poster talk code bib

 @inproceedings{qin2024label,
  title         = "A Label is Worth A Thousand Images in Dataset Distillation",
  author        = "Qin, Tian and Deng, Zhiwei and Alvarez-Melis, David",
  booktitle     = "Advances in Neural Information Processing Systems",
  volume        = "37",
  publisher     = "Curran Associates",
  year          = "2024",
  eprint        = "2406.10485",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2406.10485"
}

Tag-LLM: Repurposing General-Purpose LLMs for Specialized Domains

Junhong Shen, Neil Tenenholtz, James Brian Hall, David Alvarez-Melis, Nicolo Fusi

ICML'24: International Conference on Machine Learning. 2024.

pdf arxiv openreview poster code bib

 @inproceedings{shen2024tag-llm,
  title         = "Tag-{LLM}: Repurposing General-Purpose {LLMs} for Specialized Domains",
  author        = "Shen, Junhong and Tenenholtz, Neil and Hall, James Brian and Alvarez-Melis, David and Fusi, Nicolo",
  booktitle     = "Proceedings of the 41st International Conference on Machine Learning",
  volume        = "235",
  pages         = "44759--44773",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2024",
  eprint        = "2402.05140",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2402.05140"
}

Adapting Language Models via Token Translation

Zhili Feng, Tanya Marwah, Lester Mackey, David Alvarez-Melis, Nicolò Fusi

NeurIPS 2024 Workshop on Machine Learning and Compression.

arxiv openreview bib

 @inproceedings{feng2024adapting,
  title         = "Adapting Language Models via Token Translation",
  author        = "Feng, Zhili and Marwah, Tanya and Mackey, Lester and Alvarez-Melis, David and Fusi, Nicolò",
  booktitle     = "NeurIPS 2024 Workshop on Machine Learning and Compression",
  year          = "2024",
  eprint        = "2411.00593",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2411.00593"
}

CharED: Character-wise Ensemble Decoding for Large Language Models

Kevin Gu, Eva Tuecke, Dmitriy Katz, Raya Horesh, David Alvarez-Melis, Mikhail Yurochkin

ICML 2024 Workshop on Foundation Models in the Wild.

arxiv openreview bib

 @inproceedings{gu2024chared,
  title         = "{CharED}: Character-wise Ensemble Decoding for Large Language Models",
  author        = "Gu, Kevin and Tuecke, Eva and Katz, Dmitriy and Horesh, Raya and Alvarez-Melis, David and Yurochkin, Mikhail",
  booktitle     = "ICML 2024 Workshop on Foundation Models in the Wild",
  year          = "2024",
  eprint        = "2407.11009",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2407.11009"
}

Mixture of Parrots: Mixtures of experts improve memorization more than reasoning

Samy Jelassi, Clara Mohri, David Brandfonbrener, Alex Gu, Nikhil Vyas, Nikhil Anand, David Alvarez-Melis, Yuanzhi Li, Sham M. Kakade, Eran Malach

NeurIPS 2024 Workshop on Mathematics of Modern Machine Learning. Oral.

openreview talk bib

 @inproceedings{jelassi2024mixture,
  title     = "Mixture of Parrots: Mixtures of experts improve memorization more than reasoning",
  author    = "Jelassi, Samy and Mohri, Clara and Brandfonbrener, David and Gu, Alex and Vyas, Nikhil and Anand, Nikhil and Alvarez-Melis, David and Li, Yuanzhi and Kakade, Sham M. and Malach, Eran",
  booktitle = "NeurIPS 2024 Workshop on Mathematics of Modern Machine Learning",
  year      = "2024"
}

Distributional Dataset Distillation with Subtask Decomposition

Tian Qin, Zhiwei Deng, David Alvarez-Melis

ICLR 2024 Workshop on Navigating and Addressing Data Problems for Foundation Models.

arxiv openreview code bib

 @inproceedings{qin2024distributional,
  title         = "Distributional Dataset Distillation with Subtask Decomposition",
  author        = "Qin, Tian and Deng, Zhiwei and Alvarez-Melis, David",
  booktitle     = "ICLR 2024 Workshop on Navigating and Addressing Data Problems for Foundation Models",
  year          = "2024",
  eprint        = "2403.00999",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2403.00999"
}

Understanding the Role of Functional Diversity in Weight-Ensembling with Ingredient Selection and Multidimensional Scaling

Alex Rojas, David Alvarez-Melis

ICML 2024 Workshop on Foundation Models in the Wild.

arxiv openreview bib

 @inproceedings{rojas2024understanding,
  title         = "Understanding the Role of Functional Diversity in Weight-Ensembling with Ingredient Selection and Multidimensional Scaling",
  author        = "Rojas, Alex and Alvarez-Melis, David",
  booktitle     = "ICML 2024 Workshop on Foundation Models in the Wild",
  year          = "2024",
  eprint        = "2409.02347",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2409.02347"
}

Strongly Isomorphic Neural Optimal Transport Across Incomparable Spaces

Athina Sotiropoulou, David Alvarez-Melis

Proceedings of the Geometry-grounded Representation Learning and Generative Modeling Workshop (GRaM), 2024.

arxiv code bib

 @inproceedings{sotiropoulou2024strongly,
  title         = "Strongly Isomorphic Neural Optimal Transport Across Incomparable Spaces",
  author        = "Sotiropoulou, Athina and Alvarez-Melis, David",
  booktitle     = "Proceedings of the Geometry-grounded Representation Learning and Generative Modeling Workshop (GRaM)",
  year          = "2024",
  eprint        = "2407.14957",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2407.14957"
}

InfoOT: Information Maximizing Optimal Transport

Ching-Yao Chuang, Stefanie Jegelka, David Alvarez-Melis

ICML'23: International Conference on Machine Learning. 2023.

pdf arxiv talk code bib

 @inproceedings{chuang2023infoot,
  title         = "{InfoOT}: Information Maximizing Optimal Transport",
  author        = "Chuang, Ching-Yao and Jegelka, Stefanie and Alvarez-Melis, David",
  booktitle     = "Proceedings of the 40th International Conference on Machine Learning",
  volume        = "202",
  pages         = "6228--6242",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2023",
  eprint        = "2210.03164",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2210.03164"
}

Domain adaptation using optimal transport for invariant learning using histopathology datasets

Kianoush Falahkheirkhah, Alex Xijie Lu, David Alvarez-Melis, Grace Huynh

MIDL'23: Medical Imaging with Deep Learning. 2023.

pdf arxiv poster bib

 @inproceedings{falahkheirkhah2023domain,
  title         = "Domain adaptation using optimal transport for invariant learning using histopathology datasets",
  author        = "Falahkheirkhah, Kianoush and Lu, Alex Xijie and Alvarez-Melis, David and Huynh, Grace",
  booktitle     = "Medical Imaging with Deep Learning",
  volume        = "227",
  pages         = "1765--1782",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2023",
  eprint        = "2303.02241",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2303.02241"
}

Generating Synthetic Datasets by Interpolating along Generalized Geodesics

Jiaojiao Fan, David Alvarez-Melis

UAI'23: Uncertainty in Artificial Intelligence. 2023.

pdf arxiv openreview poster talk code bib

 @inproceedings{fan2023generating,
  title         = "Generating Synthetic Datasets by Interpolating along Generalized Geodesics",
  author        = "Fan, Jiaojiao and Alvarez-Melis, David",
  booktitle     = "Proceedings of the Thirty-Ninth Conference on Uncertainty in Artificial Intelligence",
  volume        = "216",
  pages         = "571--581",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2023",
  eprint        = "2306.06866",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2306.06866"
}

Enhancing Language Models for Technical Domains with Dynamic Token Injection

Giorgio Giannone, Neil Tenenholtz, James Hall, Nicolò Fusi, David Alvarez-Melis

NeurIPS 2023 Workshop on Generative AI and Biology.

openreview poster bib

 @inproceedings{giannone2023enhancing,
  title     = "Enhancing Language Models for Technical Domains with Dynamic Token Injection",
  author    = "Giannone, Giorgio and Tenenholtz, Neil and Hall, James and Fusi, Nicolò and Alvarez-Melis, David",
  booktitle = "NeurIPS 2023 Workshop on Generative AI and Biology",
  year      = "2023"
}

Undo Maps: A Tool for Adapting Policies to Perceptual Distortions

Abhi Gupta, Ted Moskovitz, David Alvarez-Melis, Aldo Pacchiano

ICML 2023 Workshop on New Frontiers in Learning, Control, and Dynamical Systems.

openreview bib

 @inproceedings{gupta2023undo,
  title     = "Undo Maps: A Tool for Adapting Policies to Perceptual Distortions",
  author    = "Gupta, Abhi and Moskovitz, Ted and Alvarez-Melis, David and Pacchiano, Aldo",
  booktitle = "ICML 2023 Workshop on New Frontiers in Learning, Control, and Dynamical Systems",
  year      = "2023"
}

Are GANs overkill for NLP?

David Alvarez-Melis, Vikas Garg, Adam Kalai

NeurIPS'22: Neural Information Processing Systems. 2022.

pdf arxiv openreview talk bib

 @inproceedings{alvarez-melis2022gans,
  title         = "Are {GANs} overkill for {NLP}?",
  author        = "Alvarez-Melis, David and Garg, Vikas and Kalai, Adam",
  booktitle     = "Advances in Neural Information Processing Systems",
  volume        = "35",
  pages         = "9072--9084",
  publisher     = "Curran Associates, Inc.",
  year          = "2022",
  eprint        = "2205.09838",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2205.09838"
}

Optimizing Functionals on the Space of Probabilities with Input Convex Neural Networks

David Alvarez-Melis, Yair Schiff, Youssef Mroueh

TMLR: Transactions on Machine Learning Research. 2022.

Earlier version at OTML: NeurIPS'21 Workshop on Optimal Transport in Machine Learning. Spotlight.

pdf arxiv openreview poster talk code bib

 @article{alvarez-melis2022optimizing,
  title         = "Optimizing Functionals on the Space of Probabilities with Input Convex Neural Networks",
  author        = "Alvarez-Melis, David and Schiff, Yair and Mroueh, Youssef",
  journal       = "Transactions on Machine Learning Research",
  year          = "2022",
  eprint        = "2106.00774",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2106.00774"
}

Hierarchical Optimal Transport for Comparing Histopathology Datasets

Anna Yeaton, Rahul G. Krishnan, Rebecca Mieloszyk, David Alvarez-Melis, Grace Huynh

MIDL'22: Medical Imaging with Deep Learning. 2022.

pdf arxiv talk code bib

 @inproceedings{yeaton2022hierarchical,
  title         = "Hierarchical Optimal Transport for Comparing Histopathology Datasets",
  author        = "Yeaton, Anna and Krishnan, Rahul G. and Mieloszyk, Rebecca and Alvarez-Melis, David and Huynh, Grace",
  booktitle     = "Proceedings of The 5th International Conference on Medical Imaging with Deep Learning",
  volume        = "172",
  pages         = "1459--1469",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2022",
  eprint        = "2204.08324",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2204.08324"
}

Interpretable Distribution Shift Detection using Optimal Transport

Neha Hulkund, Nicolò Fusi, Jennifer Wortman Vaughan, David Alvarez-Melis

ICML 2022 Workshop on DataPerf: Benchmarking Data for Data-Centric AI.

arxiv bib

 @inproceedings{hulkund2022interpretable,
  title         = "{Interpretable Distribution Shift Detection using Optimal Transport}",
  author        = "Hulkund, Neha and Fusi, Nicolò and Vaughan, Jennifer Wortman and Alvarez-Melis, David",
  booktitle     = "ICML 2022 Workshop on DataPerf: Benchmarking Data for Data-Centric AI",
  year          = "2022",
  eprint        = "2208.02896",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2208.02896"
}

Neural Unbalanced Optimal Transport via Cycle-Consistent Semi-Couplings

Frederike Lübeck, Charlotte Bunne, Gabriele Gut, Jacobo Sarabia del Castillo, Lucas Pelkmans, David Alvarez-Melis

NeurIPS 2022 Workshop on Learning Meaningful Representations of Life.

arxiv openreview code bib

 @inproceedings{lubeck2022neural,
  title         = "Neural Unbalanced Optimal Transport via Cycle-Consistent Semi-Couplings",
  author        = "Lübeck, Frederike and Bunne, Charlotte and Gut, Gabriele and Sarabia del Castillo, Jacobo and Pelkmans, Lucas and Alvarez-Melis, David",
  booktitle     = "NeurIPS 2022 Workshop on Learning Meaningful Representations of Life",
  year          = "2022",
  eprint        = "2209.15621",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2209.15621"
}

Budget-Constrained Bounds for Mini-Batch Estimation of Optimal Transport

David Alvarez-Melis, Nicolò Fusi, Lester Mackey, Tal Wagner

arXiv preprint. 2022.

arxiv bib

 @misc{alvarez-melis2022budget,
  title         = "Budget-Constrained Bounds for Mini-Batch Estimation of Optimal Transport",
  author        = "Alvarez-Melis, David and Fusi, Nicolò and Mackey, Lester and Wagner, Tal",
  year          = "2022",
  eprint        = "2210.13630",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2210.13630"
}

Transfer RL via the Undo Maps Formalism

Abhi Gupta, Ted Moskovitz, David Alvarez-Melis, Aldo Pacchiano

arXiv preprint. 2022.

arxiv bib

 @misc{gupta2022transfer,
  title         = "Transfer {RL} via the Undo Maps Formalism",
  author        = "Gupta, Abhi and Moskovitz, Ted and Alvarez-Melis, David and Pacchiano, Aldo",
  year          = "2022",
  eprint        = "2211.14469",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2211.14469"
}

Dataset Dynamics via Gradient Flows in Probability Space

David Alvarez-Melis, Nicolò Fusi

ICML'21: International Conference on Machine Learning. 2021.

pdf arxiv poster talk code bib

 @inproceedings{alvarez-melis2021dataset,
  title         = "Dataset Dynamics via Gradient Flows in Probability Space",
  author        = "Alvarez-Melis, David and Fusi, Nicolò",
  booktitle     = "Proceedings of the 38th International Conference on Machine Learning",
  volume        = "139",
  pages         = "219--230",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2021",
  eprint        = "2010.12760",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2010.12760"
}

From Human Explanation to Model Interpretability: A Framework Based on Weight of Evidence

David Alvarez-Melis, Harmanpreet Kaur, Hal Daumé III, Hanna Wallach, Jennifer Wortman Vaughan

HCOMP'21: AAAI Conference on Human Computation and Crowdsourcing. 2021.

pdf arxiv slides code bib

 @inproceedings{alvarez-melis2021human,
  title         = "{From Human Explanation to Model Interpretability: A Framework Based on Weight of Evidence}",
  author        = "Alvarez-Melis, David and Kaur, Harmanpreet and Daumé, III, Hal and Wallach, Hanna and Wortman Vaughan, Jennifer",
  booktitle     = "Proceedings of the Ninth AAAI Conference on Human Computation and Crowdsourcing",
  volume        = "9",
  pages         = "35--47",
  publisher     = "The AAAI Press",
  year          = "2021",
  eprint        = "2104.13299",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2104.13299"
}

Geometric Dataset Distances via Optimal Transport

David Alvarez-Melis, Nicolò Fusi

NeurIPS'20: Neural Information Processing Systems. 2020.

Earlier version at AutoML @ ICML 2020.

pdf arxiv poster talk blog code bib

 @inproceedings{alvarez-melis2020geometric,
  title         = "Geometric Dataset Distances via Optimal Transport",
  author        = "Alvarez-Melis, David and Fusi, Nicolò",
  booktitle     = "Advances in Neural Information Processing Systems",
  volume        = "33",
  pages         = "21428--21439",
  publisher     = "Curran Associates, Inc.",
  year          = "2020",
  eprint        = "2002.02923",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2002.02923"
}

Unsupervised Hierarchy Matching with Optimal Transport over Hyperbolic Spaces

David Alvarez-Melis, Youssef Mroueh, Tommi Jaakkola

AISTATS'20: International Conference on Artificial Intelligence and Statistics. 2020.

Earlier version at OTML: NeurIPS'18 Workshop on Optimal Transport for Machine Learning. Spotlight.

pdf supplement arxiv talk bib

 @inproceedings{alvarez-melis2020unsupervised,
  title         = "Unsupervised Hierarchy Matching with Optimal Transport over Hyperbolic Spaces",
  author        = "Alvarez-Melis, David and Mroueh, Youssef and Jaakkola, Tommi",
  booktitle     = "Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics",
  volume        = "108",
  pages         = "1606--1617",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2020",
  eprint        = "1911.02536",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1911.02536"
}

Towards Optimal Transport with Global Invariances

David Alvarez-Melis, Stefanie Jegelka, Tommi S. Jaakkola

AISTATS'19: International Conference on Artificial Intelligence and Statistics. 2019.

pdf supplement arxiv poster bib

 @inproceedings{alvarez-melis2019towards,
  title         = "{Towards Optimal Transport with Global Invariances}",
  author        = "Alvarez-Melis, David and Jegelka, Stefanie and Jaakkola, Tommi S.",
  booktitle     = "Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics",
  volume        = "89",
  pages         = "1870--1879",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2019",
  eprint        = "1806.09277",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1806.09277"
}

Learning Generative Models across Incomparable Spaces

Charlotte Bunne, David Alvarez-Melis, Andreas Krause, Stefanie Jegelka

ICML'19: International Conference on Machine Learning. 2019.

Earlier version at R2L: NeurIPS'18 Workshop on Relational Representation Learning. Best Paper Award.

pdf supplement arxiv slides talk code bib

 @inproceedings{bunne2019learning,
  title         = "Learning Generative Models across Incomparable Spaces",
  author        = "Bunne, Charlotte and Alvarez-Melis, David and Krause, Andreas and Jegelka, Stefanie",
  booktitle     = "Proceedings of the 36th International Conference on Machine Learning",
  volume        = "97",
  pages         = "851--861",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2019",
  eprint        = "1905.05461",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1905.05461"
}

Functional Transparency for Structured Data: a Game-Theoretic Approach

Guang-He Lee, Wengong Jin, David Alvarez-Melis, Tommi Jaakkola

ICML'19: International Conference on Machine Learning. 2019.

pdf supplement arxiv slides bib

 @inproceedings{lee2019functional,
  title         = "{Functional Transparency for Structured Data: a Game-Theoretic Approach}",
  author        = "Lee, Guang-He and Jin, Wengong and Alvarez-Melis, David and Jaakkola, Tommi",
  booktitle     = "Proceedings of the 36th International Conference on Machine Learning",
  volume        = "97",
  pages         = "3723--3733",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2019",
  eprint        = "1902.09737",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1902.09737"
}

Towards Robust, Locally Linear Deep Networks

Guang-He Lee, David Alvarez-Melis, Tommi S. Jaakkola

ICLR'19: International Conference on Learning Representations. 2019.

pdf arxiv openreview website code bib

 @inproceedings{lee2019towards,
  title         = "{Towards Robust, Locally Linear Deep Networks}",
  author        = "Lee, Guang-He and Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "International Conference on Learning Representations",
  year          = "2019",
  eprint        = "1907.03207",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1907.03207"
}

Weight of Evidence as a Basis for Human-Oriented Explanations

David Alvarez-Melis, Hal Daumé III, Jennifer Wortman Vaughan, Hanna Wallach

NeurIPS 2019 Workshop on Human-Centric Machine Learning.

arxiv code bib

 @inproceedings{alvarez-melis2019weight,
  title         = "Weight of Evidence as a Basis for Human-Oriented Explanations",
  author        = "Alvarez-Melis, David and Daumé, III, Hal and Vaughan, Jennifer Wortman and Wallach, Hanna",
  booktitle     = "NeurIPS 2019 Workshop on Human-Centric Machine Learning",
  year          = "2019",
  eprint        = "1910.13503",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1910.13503"
}

Probabilistic Bias Mitigation in Word Embeddings

Hailey Joren, David Alvarez-Melis

NeurIPS 2019 Workshop on Human-Centric Machine Learning.

arxiv bib

 @inproceedings{james-sorenson2019probabilistic,
  title         = "{Probabilistic Bias Mitigation in Word Embeddings}",
  author        = "Joren, Hailey and Alvarez-Melis, David",
  booktitle     = "NeurIPS 2019 Workshop on Human-Centric Machine Learning",
  year          = "2019",
  eprint        = "1910.14497",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1910.14497"
}

Optimal Transport in Structured Domains: Algorithms and Applications

David Alvarez-Melis (advisor: Tommi S. Jaakkola)

PhD Thesis, MIT. 2019.

pdf slides bib

 @phdthesis{alvarez-melis2019optimal,
  title  = "{Optimal Transport in Structured Domains: Algorithms and Applications}",
  author = "Alvarez-Melis, David",
  school = "Massachusetts Institute of Technology",
  year   = "2019"
}

Gromov-Wasserstein Alignment of Word Embedding Spaces

David Alvarez-Melis, Tommi Jaakkola

EMNLP'18: Empirical Methods in Natural Language Processing. 2018. Oral.

pdf arxiv slides poster talk code press bib

 @inproceedings{alvarez-melis2018gromov,
  title         = "{Gromov-Wasserstein} Alignment of Word Embedding Spaces",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi",
  booktitle     = "Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing",
  pages         = "1881--1890",
  publisher     = "Association for Computational Linguistics",
  year          = "2018",
  eprint        = "1809.00013",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1809.00013"
}

Towards Robust Interpretability with Self-Explaining Neural Networks

David Alvarez-Melis, Tommi S. Jaakkola

NeurIPS'18: Neural Information Processing Systems. 2018.

pdf arxiv poster code bib

 @inproceedings{alvarez-melis2018self-explaining,
  title         = "Towards Robust Interpretability with Self-Explaining Neural Networks",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "Advances in Neural Information Processing Systems",
  volume        = "31",
  pages         = "7775--7784",
  publisher     = "Curran Associates, Inc.",
  year          = "2018",
  eprint        = "1806.07538",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1806.07538"
}

Structured Optimal Transport

David Alvarez-Melis, Tommi Jaakkola, Stefanie Jegelka

AISTATS'18: International Conference on Artificial Intelligence and Statistics. 2018. Oral.

Earlier version at NIPS Workshop on Optimal Transport for Machine Learning, 2017, as Extended Oral.

pdf supplement arxiv poster bib

 @inproceedings{alvarez-melis2018structured,
  title         = "Structured Optimal Transport",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi and Jegelka, Stefanie",
  booktitle     = "Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics",
  volume        = "84",
  pages         = "1771--1780",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2018",
  eprint        = "1712.06199",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1712.06199"
}

On the Robustness of Interpretability Methods

David Alvarez-Melis, Tommi S. Jaakkola

ICML 2018 Workshop on Human Interpretability in Machine Learning.

arxiv slides code bib

 @inproceedings{alvarez-melis2018robustness,
  title         = "{On the Robustness of Interpretability Methods}",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "ICML 2018 Workshop on Human Interpretability in Machine Learning",
  year          = "2018",
  eprint        = "1806.08049",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1806.08049"
}

Learning Generative Models across Incomparable Spaces

Charlotte Bunne, David Alvarez-Melis, Andreas Krause, Stefanie Jegelka

NeurIPS 2018 Workshop on Relational Representation Learning. Best Paper Award.

pdf code bib

 @inproceedings{bunne2018learning,
  title     = "{Learning Generative Models across Incomparable Spaces}",
  author    = "Bunne, Charlotte and Alvarez-Melis, David and Krause, Andreas and Jegelka, Stefanie",
  booktitle = "NeurIPS 2018 Workshop on Relational Representation Learning",
  year      = "2018"
}

Game-Theoretic Interpretability for Temporal Modeling

Guang-He Lee, David Alvarez-Melis, Tommi S. Jaakkola

ICML 2018 Workshop on Fairness, Accountability, and Transparency in Machine Learning.

arxiv bib

 @inproceedings{lee2018game-theoretic,
  title         = "{Game-Theoretic Interpretability for Temporal Modeling}",
  author        = "Lee, Guang-He and Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "ICML 2018 Workshop on Fairness, Accountability, and Transparency in Machine Learning",
  year          = "2018",
  eprint        = "1807.00130",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1807.00130"
}

Distributional Adversarial Networks

Chengtao Li, David Alvarez-Melis, Keyulu Xu, Stefanie Jegelka, Suvrit Sra

ICLR 2018 Workshop Track.

arxiv openreview code bib

 @inproceedings{li2018distributional,
  title         = "Distributional Adversarial Networks",
  author        = "Li, Chengtao and Alvarez-Melis, David and Xu, Keyulu and Jegelka, Stefanie and Sra, Suvrit",
  booktitle     = "ICLR 2018 Workshop Track",
  year          = "2018",
  eprint        = "1706.09549",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1706.09549"
}

A causal framework for explaining the predictions of black-box sequence-to-sequence models

David Alvarez-Melis, Tommi S. Jaakkola

EMNLP'17: Empirical Methods in Natural Language Processing. 2017.

pdf supplement arxiv press bib

 @inproceedings{alvarez-melis2017causal,
  title         = "A causal framework for explaining the predictions of black-box sequence-to-sequence models",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing",
  pages         = "412--421",
  publisher     = "Association for Computational Linguistics",
  year          = "2017",
  eprint        = "1707.01943",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1707.01943"
}

Tree-structured decoding with doubly-recurrent neural networks

David Alvarez-Melis, Tommi S. Jaakkola

ICLR'17: International Conference on Learning Representations. 2017.

pdf openreview poster code bib

 @inproceedings{alvarez-melis2017tree,
  title     = "Tree-structured decoding with doubly-recurrent neural networks",
  author    = "Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle = "International Conference on Learning Representations",
  year      = "2017"
}

The Emotional GAN: Priming Adversarial Generation of Art with Emotion

David Alvarez-Melis, Judith Amores

NIPS 2017 Workshop on Machine Learning for Creativity and Design.

pdf bib

 @inproceedings{alvarez-melis2017emotional,
  title     = "{The Emotional GAN: Priming Adversarial Generation of Art with Emotion}",
  author    = "Alvarez-Melis, David and Amores, Judith",
  booktitle = "NIPS 2017 Workshop on Machine Learning for Creativity and Design",
  year      = "2017"
}

Topic Modeling in Twitter: Aggregating Tweets by Conversations

David Alvarez-Melis, Martin Saveski

ICWSM'16: International AAAI Conference on Web and Social Media. 2016. (Short paper.)

pdf poster bib

 @inproceedings{alvarez-melis2016topic,
  title     = "Topic Modeling in {Twitter}: Aggregating Tweets by Conversations",
  author    = "Alvarez-Melis, David and Saveski, Martin",
  booktitle = "Proceedings of the International AAAI Conference on Web and Social Media",
  volume    = "10",
  pages     = "519--522",
  publisher = "AAAI",
  year      = "2016"
}

Word Embeddings as Metric Recovery in Semantic Spaces

Tatsunori B. Hashimoto, David Alvarez-Melis, Tommi S. Jaakkola

TACL: Transactions of the Association for Computational Linguistics. 2016. (Presented at ACL'16.)

pdf bib

 @article{hashimoto2016word,
  title   = "{Word Embeddings as Metric Recovery in Semantic Spaces}",
  author  = "Hashimoto, Tatsunori B. and Alvarez-Melis, David and Jaakkola, Tommi S.",
  journal = "Transactions of the Association for Computational Linguistics",
  volume  = "4",
  pages   = "273--286",
  doi     = "10.1162/tacl_a_00098",
  year    = "2016"
}

Word, graph and manifold embedding from Markov processes

Tatsunori B. Hashimoto, David Alvarez-Melis, Tommi S. Jaakkola

NIPS 2015 Workshop on Nonparametric Methods for Large Scale Representation Learning. Oral.

pdf arxiv bib

 @inproceedings{hashimoto2015word,
  title         = "{Word, graph and manifold embedding from Markov processes}",
  author        = "Hashimoto, Tatsunori B. and Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "NIPS 2015 Workshop on Nonparametric Methods for Large Scale Representation Learning",
  year          = "2015",
  eprint        = "1509.05808",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1509.05808"
}

A translation of “The characteristic function of a random phenomenon” by Bruno de Finetti

David Alvarez-Melis, Tamara Broderick

arXiv preprint. 2015.

arxiv bib

 @misc{alvarez-melis2015translation,
  title         = "A translation of ``The characteristic function of a random phenomenon'' by Bruno de Finetti",
  author        = "Alvarez-Melis, David and Broderick, Tamara",
  year          = "2015",
  eprint        = "1512.01229",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1512.01229"
}

The Matrix Multiplicative Weights Algorithm for Domain Adaptation

David Alvarez-Melis (advisor: Mehryar Mohri)

MS Thesis, Courant Institute, NYU. 2013.

pdf bib

 @mastersthesis{alvarez-melis2013matrix,
  title  = "{The Matrix Multiplicative Weights Algorithm for Domain Adaptation}",
  author = "Alvarez-Melis, David",
  school = "New York University",
  year   = "2013"
}

Lax-Milgram's Theorem: Generalizations and Applications

David Alvarez-Melis (advisor: Carlos Bosch Giral)

BSc Thesis, ITAM. 2011.

pdf bib

 @misc{alvarez-melis2011lax,
  title        = "{El Teorema de Lax-Milgram, Generalizaciones y Aplicaciones}",
  author       = "Alvarez-Melis, David",
  school       = "Instituto Tecnológico Autónomo de México",
  howpublished = "BSc thesis",
  year         = "2011"
}

Bridging Compute- and Data-Optimal Pretraining

Tian Qin, Kimia Hamidieh, David Alvarez-Melis

NeurIPS'26: Neural Information Processing Systems. 2026. To appear. Oral.

pdf arxiv code bib

 @inproceedings{qin2026bridging,
  title         = "Bridging Compute- and Data-Optimal Pretraining",
  author        = "Qin, Tian and Hamidieh, Kimia and Alvarez-Melis, David",
  booktitle     = "Advances in Neural Information Processing Systems",
  year          = "2026",
  eprint        = "2607.25271",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2607.25271"
}

Domain-Aware Scaling Laws Uncover Data Synergy

Kimia Hamidieh, Lester Mackey, David Alvarez-Melis

COLM'26: Conference on Language Modeling. 2026.

arxiv openreview bib

 @inproceedings{hamidieh2026domain-aware,
  title         = "Domain-Aware Scaling Laws Uncover Data Synergy",
  author        = "Hamidieh, Kimia and Mackey, Lester and Alvarez-Melis, David",
  booktitle     = "Conference on Language Modeling",
  year          = "2026",
  eprint        = "2607.11052",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2607.11052"
}

Boomerang Distillation Enables Zero-Shot Model Size Interpolation

Sara Kangaslahti, Nihal V. Nayak, Jonathan Geuter, Marco Fumero, Francesco Locatello, David Alvarez-Melis

ICLR'26: International Conference on Learning Representations. 2026.

arxiv openreview poster blog code bib

 @inproceedings{kangaslahti2026boomerang,
  title         = "Boomerang Distillation Enables Zero-Shot Model Size Interpolation",
  author        = "Kangaslahti, Sara and Nayak, Nihal V. and Geuter, Jonathan and Fumero, Marco and Locatello, Francesco and Alvarez-Melis, David",
  booktitle     = "International Conference on Learning Representations",
  year          = "2026",
  eprint        = "2510.05064",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2510.05064"
}

A Critical Look at Targeted Instruction Selection: Disentangling What Matters (and What Doesn't)

Nihal V. Nayak, Paula Rodriguez-Diaz, Neha Hulkund, Sara Beery, David Alvarez-Melis

ICML'26: International Conference on Machine Learning. 2026.

arxiv openreview code bib

 @inproceedings{nayak2026critical,
  title         = "A Critical Look at Targeted Instruction Selection: Disentangling What Matters (and What Doesn't)",
  author        = "Nayak, Nihal V. and Rodriguez-Diaz, Paula and Hulkund, Neha and Beery, Sara and Alvarez-Melis, David",
  booktitle     = "Proceedings of the 43rd International Conference on Machine Learning",
  volume        = "306",
  pages         = "92008--92043",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2026",
  eprint        = "2602.14696",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2602.14696"
}

Curriculum Learning as Transport: Understanding Curricula with Wasserstein Geodesics

Changho Shin, David Alvarez-Melis

COLM'26: Conference on Language Modeling. 2026.

arxiv openreview bib

 @inproceedings{shin2026curriculum,
  title         = "Curriculum Learning as Transport: Understanding Curricula with Wasserstein Geodesics",
  author        = "Shin, Changho and Alvarez-Melis, David",
  booktitle     = "Conference on Language Modeling",
  year          = "2026",
  eprint        = "2609.09099",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2609.09099"
}

A Label is Worth A Thousand Images in Dataset Distillation

Tian Qin, Zhiwei Deng, David Alvarez-Melis

NeurIPS'24: Neural Information Processing Systems. 2024.

pdf arxiv openreview poster talk code bib

 @inproceedings{qin2024label,
  title         = "A Label is Worth A Thousand Images in Dataset Distillation",
  author        = "Qin, Tian and Deng, Zhiwei and Alvarez-Melis, David",
  booktitle     = "Advances in Neural Information Processing Systems",
  volume        = "37",
  publisher     = "Curran Associates",
  year          = "2024",
  eprint        = "2406.10485",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2406.10485"
}

Dataset Dynamics via Gradient Flows in Probability Space

David Alvarez-Melis, Nicolò Fusi

ICML'21: International Conference on Machine Learning. 2021.

pdf arxiv poster talk code bib

 @inproceedings{alvarez-melis2021dataset,
  title         = "Dataset Dynamics via Gradient Flows in Probability Space",
  author        = "Alvarez-Melis, David and Fusi, Nicolò",
  booktitle     = "Proceedings of the 38th International Conference on Machine Learning",
  volume        = "139",
  pages         = "219--230",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2021",
  eprint        = "2010.12760",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2010.12760"
}

Geometric Dataset Distances via Optimal Transport

David Alvarez-Melis, Nicolò Fusi

NeurIPS'20: Neural Information Processing Systems. 2020.

Earlier version at AutoML @ ICML 2020.

pdf arxiv poster talk blog code bib

 @inproceedings{alvarez-melis2020geometric,
  title         = "Geometric Dataset Distances via Optimal Transport",
  author        = "Alvarez-Melis, David and Fusi, Nicolò",
  booktitle     = "Advances in Neural Information Processing Systems",
  volume        = "33",
  pages         = "21428--21439",
  publisher     = "Curran Associates, Inc.",
  year          = "2020",
  eprint        = "2002.02923",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2002.02923"
}

Gromov-Wasserstein Alignment of Word Embedding Spaces

David Alvarez-Melis, Tommi Jaakkola

EMNLP'18: Empirical Methods in Natural Language Processing. 2018. Oral.

pdf arxiv slides poster talk code press bib

 @inproceedings{alvarez-melis2018gromov,
  title         = "{Gromov-Wasserstein} Alignment of Word Embedding Spaces",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi",
  booktitle     = "Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing",
  pages         = "1881--1890",
  publisher     = "Association for Computational Linguistics",
  year          = "2018",
  eprint        = "1809.00013",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1809.00013"
}

Towards Robust Interpretability with Self-Explaining Neural Networks

David Alvarez-Melis, Tommi S. Jaakkola

NeurIPS'18: Neural Information Processing Systems. 2018.

pdf arxiv poster code bib

 @inproceedings{alvarez-melis2018self-explaining,
  title         = "Towards Robust Interpretability with Self-Explaining Neural Networks",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "Advances in Neural Information Processing Systems",
  volume        = "31",
  pages         = "7775--7784",
  publisher     = "Curran Associates, Inc.",
  year          = "2018",
  eprint        = "1806.07538",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1806.07538"
}

Beyond Explainable AI (XAI): An Overdue Paradigm Shift and Post-XAI Research Directions

arXiv preprint. 2026.

arxiv bib

 @misc{afroogh2026beyond,
  title         = "Beyond Explainable AI (XAI): An Overdue Paradigm Shift and Post-XAI Research Directions",
  author        = "Afroogh, Saleh and Ahmed, Syed Ishtiaque and Ahrweiler, Petra and Alvarez-Melis, David and Arief, Mansur Maturidi and Barakova, Emilia and Bargagli-Stoffi, Falco J. and Biyik, Erdem and Chen, Hanjie and Chen, Xiang 'Anthony' and Clements, Robert Alan and Crockett, Keeley and Dhurandhar, Amit and Dogan, Fethiye Irmak and Dollinger, Mollie and Eslami, Motahhare and Faisal, Aldo A and Farahi, Arya and Pradier, Melanie F. and Gabriel, Saadia and Garcia-Olano, Diego and Ghassemi, Marzyeh and Ghosh, Shaona and Gunes, Hatice and Hajiramezanali, Ehsan and Haufe, Stefan and Huang, Biwei and Hwang, Angel and Islam, Md Tauhidul and Jiao, Junfeng and Karimi, Amir-Hossein and Kazeminasab, Saber and Kuzminykh, Anastasia and La Cava, William and Lim, Brian Y. and Liu, Xiaofeng and Mofrad, Mohammad R. K. and Parrish, Alicia and Perez-Ortiz, Maria and Raj, Shriti and Swayamdipta, Swabha and Talebi, Salmonn and Varshney, Kush R. and Vorvoreanu, Mihaela and Weng, Lily and Xiang, Alice and Xu, Yiming and Zhao, Ding and Zhao, Jieyu",
  year          = "2026",
  eprint        = "2602.24176",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2602.24176"
}

Releasing the Inner Jev: A Simple Recipe for Open Decision Models

Jonathan Geuter, Andrew Cheng, David Alvarez-Melis

arXiv preprint. 2026.

arxiv code dataset bib

 @misc{geuter2026releasing,
  title         = "Releasing the Inner Jev: A Simple Recipe for Open Decision Models",
  author        = "Geuter, Jonathan and Cheng, Andrew and Alvarez-Melis, David",
  year          = "2026",
  eprint        = "2610.17205",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2610.17205"
}

Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention

Jing Huang, Daniel Wurgaft, Rachit Bansal, Laura Ruis, Naomi Saphra, David Alvarez-Melis, Andrew Kyle Lampinen, Christopher Potts, Ekdeep Singh Lubana

arXiv preprint. 2026.

arxiv bib

 @misc{huang2026why,
  title         = "Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention",
  author        = "Huang, Jing and Wurgaft, Daniel and Bansal, Rachit and Ruis, Laura and Saphra, Naomi and Alvarez-Melis, David and Lampinen, Andrew Kyle and Potts, Christopher and Lubana, Ekdeep Singh",
  year          = "2026",
  eprint        = "2605.29548",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2605.29548"
}

Adversarial Concept Search: Predicting Compositional Errors From Feature Geometry

Jennifer Meng Lu, Ruochen Zhang, Isabelle Lee, David Alvarez-Melis, Ellie Pavlick, Naomi Saphra

arXiv preprint. 2026.

arxiv bib

 @misc{lu2026adversarial,
  title         = "Adversarial Concept Search: Predicting Compositional Errors From Feature Geometry",
  author        = "Lu, Jennifer Meng and Zhang, Ruochen and Lee, Isabelle and Alvarez-Melis, David and Pavlick, Ellie and Saphra, Naomi",
  year          = "2026",
  eprint        = "2606.13934",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2606.13934"
}

Universal Bounds for Self-Normalized Kullback-Leibler Estimation

Buu Phan, Lester Mackey, David Alvarez-Melis

arXiv preprint. 2026.

arxiv bib

 @misc{phan2026universal,
  title         = "Universal Bounds for Self-Normalized Kullback-Leibler Estimation",
  author        = "Phan, Buu and Mackey, Lester and Alvarez-Melis, David",
  year          = "2026",
  eprint        = "2610.12851",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2610.12851"
}

Low-Frequency Shortcuts in Texture-Driven Visual Learning

Utku Şirin, Cathy Hou, Despina-Ekaterini Argiropoulos, David Alvarez-Melis, Stratos Idreos

arXiv preprint. 2026.

arxiv bib

 @misc{sirin2026low-frequency,
  title         = "Low-Frequency Shortcuts in Texture-Driven Visual Learning",
  author        = "Şirin, Utku and Hou, Cathy and Argiropoulos, Despina-Ekaterini and Alvarez-Melis, David and Idreos, Stratos",
  year          = "2026",
  eprint        = "2606.03493",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2606.03493"
}

OT on the Map: Quantifying Domain Shifts in Geographic Space

Haoran Zhang, Livia Betti, Konstantin Klemmer, Esther Rolf, David Alvarez-Melis

arXiv preprint. 2026.

arxiv code bib

 @misc{zhang2026ot,
  title         = "OT on the Map: Quantifying Domain Shifts in Geographic Space",
  author        = "Zhang, Haoran and Betti, Livia and Klemmer, Konstantin and Rolf, Esther and Alvarez-Melis, David",
  year          = "2026",
  eprint        = "2604.16220",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2604.16220"
}

Adam or Gauss-Newton? A Comparative Study In Terms of Basis Alignment and SGD Noise

Bingbin Liu, Rachit Bansal, Depen Morwani, Nikhil Vyas, David Alvarez-Melis, Sham M. Kakade

arXiv preprint. 2025.

arxiv bib

 @misc{liu2025adam,
  title         = "Adam or Gauss-Newton? A Comparative Study In Terms of Basis Alignment and SGD Noise",
  author        = "Liu, Bingbin and Bansal, Rachit and Morwani, Depen and Vyas, Nikhil and Alvarez-Melis, David and Kakade, Sham M.",
  year          = "2025",
  eprint        = "2510.13680",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2510.13680"
}

Decomposing Elements of Problem Solving: What “Math” Does RL Teach?

Tian Qin, Core Francisco Park, Mujin Kwun, Aaron Walsman, Eran Malach, Nikhil Anand, Hidenori Tanaka, David Alvarez-Melis

arXiv preprint. 2025.

arxiv code bib

 @misc{qin2025decomposing,
  title         = "Decomposing Elements of Problem Solving: What ``Math'' Does {RL} Teach?",
  author        = "Qin, Tian and Park, Core Francisco and Kwun, Mujin and Walsman, Aaron and Malach, Eran and Anand, Nikhil and Tanaka, Hidenori and Alvarez-Melis, David",
  year          = "2025",
  eprint        = "2505.22756",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2505.22756"
}

KerJEPA: Kernel Discrepancies for Euclidean Self-Supervised Learning

Eric Zimmermann, Harley Wiltzer, Justin Szeto, David Alvarez-Melis, Lester Mackey

arXiv preprint. 2025.

arxiv bib

 @misc{zimmermann2025kerjepa,
  title         = "KerJEPA: Kernel Discrepancies for Euclidean Self-Supervised Learning",
  author        = "Zimmermann, Eric and Wiltzer, Harley and Szeto, Justin and Alvarez-Melis, David and Mackey, Lester",
  year          = "2025",
  eprint        = "2512.19605",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2512.19605"
}

Budget-Constrained Bounds for Mini-Batch Estimation of Optimal Transport

David Alvarez-Melis, Nicolò Fusi, Lester Mackey, Tal Wagner

arXiv preprint. 2022.

arxiv bib

 @misc{alvarez-melis2022budget,
  title         = "Budget-Constrained Bounds for Mini-Batch Estimation of Optimal Transport",
  author        = "Alvarez-Melis, David and Fusi, Nicolò and Mackey, Lester and Wagner, Tal",
  year          = "2022",
  eprint        = "2210.13630",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2210.13630"
}

Transfer RL via the Undo Maps Formalism

Abhi Gupta, Ted Moskovitz, David Alvarez-Melis, Aldo Pacchiano

arXiv preprint. 2022.

arxiv bib

 @misc{gupta2022transfer,
  title         = "Transfer {RL} via the Undo Maps Formalism",
  author        = "Gupta, Abhi and Moskovitz, Ted and Alvarez-Melis, David and Pacchiano, Aldo",
  year          = "2022",
  eprint        = "2211.14469",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2211.14469"
}

A translation of “The characteristic function of a random phenomenon” by Bruno de Finetti

David Alvarez-Melis, Tamara Broderick

arXiv preprint. 2015.

arxiv bib

 @misc{alvarez-melis2015translation,
  title         = "A translation of ``The characteristic function of a random phenomenon'' by Bruno de Finetti",
  author        = "Alvarez-Melis, David and Broderick, Tamara",
  year          = "2015",
  eprint        = "1512.01229",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1512.01229"
}

Bridging Compute- and Data-Optimal Pretraining

Tian Qin, Kimia Hamidieh, David Alvarez-Melis

NeurIPS'26: Neural Information Processing Systems. 2026. To appear. Oral.

pdf arxiv code bib

 @inproceedings{qin2026bridging,
  title         = "Bridging Compute- and Data-Optimal Pretraining",
  author        = "Qin, Tian and Hamidieh, Kimia and Alvarez-Melis, David",
  booktitle     = "Advances in Neural Information Processing Systems",
  year          = "2026",
  eprint        = "2607.25271",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2607.25271"
}

Let's (not) just put things in Context: Test-time Training for Long-context LLMs

ICLR'26: International Conference on Learning Representations. 2026.

arxiv openreview bib

 @inproceedings{bansal2026lets,
  title         = "Let's (not) just put things in Context: Test-time Training for Long-context {LLMs}",
  author        = "Bansal, Rachit and Zhang, Aston and Tiwari, Rishabh and Madaan, Lovish and Duvvuri, Venkata Sai Surya Subramanyam and Khatri, Devvrit and Brandfonbrener, David and Alvarez-Melis, David and Bhargava, Prajjwal and Kale, Mihir and Jelassi, Samy",
  booktitle     = "International Conference on Learning Representations",
  year          = "2026",
  eprint        = "2512.13898",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2512.13898"
}

RL Excursions during Pre-training: How early is too early for on-policy learning?

Rachit Bansal, Clara Mohri, Tian Qin, David Alvarez-Melis, Sham Kakade

NeurIPS'26: Neural Information Processing Systems. 2026. To appear.

arxiv project bib

 @inproceedings{bansal2026rl,
  title         = "{RL} Excursions during Pre-training: How early is too early for on-policy learning?",
  author        = "Bansal, Rachit and Mohri, Clara and Qin, Tian and Alvarez-Melis, David and Kakade, Sham",
  booktitle     = "Advances in Neural Information Processing Systems",
  year          = "2026",
  eprint        = "2606.04272",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2606.04272"
}

Guided Speculative Inference for Efficient Test-Time Alignment of LLMs

Jonathan Geuter, Youssef Mroueh, David Alvarez-Melis

ICLR'26: International Conference on Learning Representations. 2026.

arxiv openreview code bib

 @inproceedings{geuter2025guided,
  title         = "Guided Speculative Inference for Efficient Test-Time Alignment of {LLMs}",
  author        = "Geuter, Jonathan and Mroueh, Youssef and Alvarez-Melis, David",
  booktitle     = "International Conference on Learning Representations",
  year          = "2026",
  eprint        = "2506.04118",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2506.04118"
}

Domain-Aware Scaling Laws Uncover Data Synergy

Kimia Hamidieh, Lester Mackey, David Alvarez-Melis

COLM'26: Conference on Language Modeling. 2026.

arxiv openreview bib

 @inproceedings{hamidieh2026domain-aware,
  title         = "Domain-Aware Scaling Laws Uncover Data Synergy",
  author        = "Hamidieh, Kimia and Mackey, Lester and Alvarez-Melis, David",
  booktitle     = "Conference on Language Modeling",
  year          = "2026",
  eprint        = "2607.11052",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2607.11052"
}

Boomerang Distillation Enables Zero-Shot Model Size Interpolation

Sara Kangaslahti, Nihal V. Nayak, Jonathan Geuter, Marco Fumero, Francesco Locatello, David Alvarez-Melis

ICLR'26: International Conference on Learning Representations. 2026.

arxiv openreview poster blog code bib

 @inproceedings{kangaslahti2026boomerang,
  title         = "Boomerang Distillation Enables Zero-Shot Model Size Interpolation",
  author        = "Kangaslahti, Sara and Nayak, Nihal V. and Geuter, Jonathan and Fumero, Marco and Locatello, Francesco and Alvarez-Melis, David",
  booktitle     = "International Conference on Learning Representations",
  year          = "2026",
  eprint        = "2510.05064",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2510.05064"
}

Understanding Layer Patching in Model Size Interpolation

Sara Kangaslahti, Jonathan Geuter, Nihal V. Nayak, Marco Fumero, Francesco Locatello, David Alvarez-Melis

NeurIPS'26: Neural Information Processing Systems. 2026. To appear.

arxiv bib

 @inproceedings{kangaslahti2026layer,
  title         = "Understanding Layer Patching in Model Size Interpolation",
  author        = "Kangaslahti, Sara and Geuter, Jonathan and Nayak, Nihal V. and Fumero, Marco and Locatello, Francesco and Alvarez-Melis, David",
  booktitle     = "Advances in Neural Information Processing Systems",
  year          = "2026",
  eprint        = "2607.08170",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2607.08170"
}

Stop Training for the Worst: Progressive Unmasking Accelerates Masked Diffusion Training

Jaeyeon Kim, Jonathan Geuter, David Alvarez-Melis, Sham M. Kakade, Sitan Chen

ICML'26: International Conference on Machine Learning. 2026.

arxiv openreview code bib

 @inproceedings{kim2026stop,
  title         = "Stop Training for the Worst: Progressive Unmasking Accelerates Masked Diffusion Training",
  author        = "Kim, Jaeyeon and Geuter, Jonathan and Alvarez-Melis, David and Kakade, Sham M. and Chen, Sitan",
  booktitle     = "Proceedings of the 43rd International Conference on Machine Learning",
  volume        = "306",
  pages         = "57515--57538",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2026",
  eprint        = "2602.10314",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2602.10314"
}

Three-operator Splitting with Stale Gradients for Faster Non-linear Optimal Transport

Jacob Lindbäck, David Alvarez-Melis, Mikael Johansson

AISTATS'26: International Conference on Artificial Intelligence and Statistics. 2026.

openreview poster bib

 @inproceedings{lindback2026three-operator,
  title     = "Three-operator Splitting with Stale Gradients for Faster Non-linear Optimal Transport",
  author    = "Lindbäck, Jacob and Alvarez-Melis, David and Johansson, Mikael",
  booktitle = "Proceedings of The 29th International Conference on Artificial Intelligence and Statistics",
  volume    = "300",
  publisher = "PMLR",
  series    = "Proceedings of Machine Learning Research",
  year      = "2026"
}

A Critical Look at Targeted Instruction Selection: Disentangling What Matters (and What Doesn't)

Nihal V. Nayak, Paula Rodriguez-Diaz, Neha Hulkund, Sara Beery, David Alvarez-Melis

ICML'26: International Conference on Machine Learning. 2026.

arxiv openreview code bib

 @inproceedings{nayak2026critical,
  title         = "A Critical Look at Targeted Instruction Selection: Disentangling What Matters (and What Doesn't)",
  author        = "Nayak, Nihal V. and Rodriguez-Diaz, Paula and Hulkund, Neha and Beery, Sara and Alvarez-Melis, David",
  booktitle     = "Proceedings of the 43rd International Conference on Machine Learning",
  volume        = "306",
  pages         = "92008--92043",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2026",
  eprint        = "2602.14696",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2602.14696"
}

Curriculum Learning as Transport: Understanding Curricula with Wasserstein Geodesics

Changho Shin, David Alvarez-Melis

COLM'26: Conference on Language Modeling. 2026.

arxiv openreview bib

 @inproceedings{shin2026curriculum,
  title         = "Curriculum Learning as Transport: Understanding Curricula with Wasserstein Geodesics",
  author        = "Shin, Changho and Alvarez-Melis, David",
  booktitle     = "Conference on Language Modeling",
  year          = "2026",
  eprint        = "2609.09099",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2609.09099"
}

Integrating chemical structures as treatments improves representations of microscopy images for morphological profiling

Yemin Yu, Emre Hayir, Neil Tenenholtz, Lester Mackey, Ying Wei, David Alvarez-Melis, Ava P. Amini, Alex X. Lu

PLOS Computational Biology. 2026.

arxiv code dataset bib

 @article{yu2025causal,
  title         = "Integrating chemical structures as treatments improves representations of microscopy images for morphological profiling",
  author        = "Yu, Yemin and Hayir, Emre and Tenenholtz, Neil and Mackey, Lester and Wei, Ying and Alvarez-Melis, David and Amini, Ava P. and Lu, Alex X.",
  journal       = "PLOS Computational Biology",
  volume        = "22",
  number        = "7",
  doi           = "10.1371/journal.pcbi.1014483",
  year          = "2026",
  eprint        = "2504.09544",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2504.09544"
}

Random Scaling of Emergent Capabilities

Rosie Zhao, Tian Qin, David Alvarez-Melis, Sham M. Kakade, Naomi Saphra

ICML'26: International Conference on Machine Learning. 2026.

arxiv openreview slides bib

 @inproceedings{zhao2025distributional,
  title         = "Random Scaling of Emergent Capabilities",
  author        = "Zhao, Rosie and Qin, Tian and Alvarez-Melis, David and Kakade, Sham M. and Saphra, Naomi",
  booktitle     = "Proceedings of the 43rd International Conference on Machine Learning",
  volume        = "306",
  pages         = "161980--161999",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2026",
  eprint        = "2502.17356",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2502.17356"
}

Thinking at the Right Size: Amortized Distillation Across Post-Trained LLMs

Yan Zhou, Sara Kangaslahti, Jonathan Geuter, Nihal V. Nayak, Marco Fumero, Francesco Locatello, David Alvarez-Melis

EMNLP Findings'26: Findings of the Association for Computational Linguistics: EMNLP. 2026.

arxiv code bib

 @inproceedings{zhou2026thinking,
  title         = "Thinking at the Right Size: Amortized Distillation Across Post-Trained {LLMs}",
  author        = "Zhou, Yan and Kangaslahti, Sara and Geuter, Jonathan and Nayak, Nihal V. and Fumero, Marco and Locatello, Francesco and Alvarez-Melis, David",
  booktitle     = "Findings of the Association for Computational Linguistics: EMNLP 2026",
  year          = "2026",
  eprint        = "2608.22854",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2608.22854"
}

Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention

Jing Huang, Daniel Wurgaft, Rachit Bansal, Laura Ruis, Naomi Saphra, David Alvarez-Melis, Andrew Kyle Lampinen, Christopher Potts, Ekdeep Singh Lubana

ICML 2026 Workshop on High Dimensional Learning Dynamics.

arxiv bib

 @inproceedings{huang2026why-hild,
  title         = "Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention",
  author        = "Huang, Jing and Wurgaft, Daniel and Bansal, Rachit and Ruis, Laura and Saphra, Naomi and Alvarez-Melis, David and Lampinen, Andrew Kyle and Potts, Christopher and Lubana, Ekdeep Singh",
  booktitle     = "ICML 2026 Workshop on High Dimensional Learning Dynamics",
  year          = "2026",
  eprint        = "2605.29548",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2605.29548"
}

Adam or Gauss-Newton? A Comparative Study In Terms of Basis Alignment and SGD Noise

Bingbin Liu, Rachit Bansal, Depen Morwani, David Alvarez-Melis, Sham M. Kakade

ICML 2026 Workshop on High Dimensional Learning Dynamics.

arxiv openreview bib

 @inproceedings{liu2026adam-hild,
  title         = "Adam or Gauss-Newton? A Comparative Study In Terms of Basis Alignment and SGD Noise",
  author        = "Liu, Bingbin and Bansal, Rachit and Morwani, Depen and Alvarez-Melis, David and Kakade, Sham M.",
  booktitle     = "ICML 2026 Workshop on High Dimensional Learning Dynamics",
  year          = "2026",
  eprint        = "2510.13680",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2510.13680"
}

Adam or Gauss-Newton? A Comparative Study In Terms of Basis Alignment and SGD Noise

Bingbin Liu, Rachit Bansal, Depen Morwani, Nikhil Vyas, David Alvarez-Melis, Sham M. Kakade

NeurIPS 2026 Workshop on Optimization for Machine Learning. To appear.

arxiv openreview bib

 @inproceedings{liu2026adam-opt,
  title         = "Adam or Gauss-Newton? A Comparative Study In Terms of Basis Alignment and SGD Noise",
  author        = "Liu, Bingbin and Bansal, Rachit and Morwani, Depen and Vyas, Nikhil and Alvarez-Melis, David and Kakade, Sham M.",
  booktitle     = "NeurIPS 2026 Workshop on Optimization for Machine Learning",
  year          = "2026",
  eprint        = "2510.13680",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2510.13680"
}

Releasing the Inner Jev: A Simple Recipe for Open Decision Models

Jonathan Geuter, Andrew Cheng, David Alvarez-Melis

arXiv preprint. 2026.

arxiv code dataset bib

 @misc{geuter2026releasing,
  title         = "Releasing the Inner Jev: A Simple Recipe for Open Decision Models",
  author        = "Geuter, Jonathan and Cheng, Andrew and Alvarez-Melis, David",
  year          = "2026",
  eprint        = "2610.17205",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2610.17205"
}

Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention

Jing Huang, Daniel Wurgaft, Rachit Bansal, Laura Ruis, Naomi Saphra, David Alvarez-Melis, Andrew Kyle Lampinen, Christopher Potts, Ekdeep Singh Lubana

arXiv preprint. 2026.

arxiv bib

 @misc{huang2026why,
  title         = "Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention",
  author        = "Huang, Jing and Wurgaft, Daniel and Bansal, Rachit and Ruis, Laura and Saphra, Naomi and Alvarez-Melis, David and Lampinen, Andrew Kyle and Potts, Christopher and Lubana, Ekdeep Singh",
  year          = "2026",
  eprint        = "2605.29548",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2605.29548"
}

Universal Bounds for Self-Normalized Kullback-Leibler Estimation

Buu Phan, Lester Mackey, David Alvarez-Melis

arXiv preprint. 2026.

arxiv bib

 @misc{phan2026universal,
  title         = "Universal Bounds for Self-Normalized Kullback-Leibler Estimation",
  author        = "Phan, Buu and Mackey, Lester and Alvarez-Melis, David",
  year          = "2026",
  eprint        = "2610.12851",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2610.12851"
}

Low-Frequency Shortcuts in Texture-Driven Visual Learning

Utku Şirin, Cathy Hou, Despina-Ekaterini Argiropoulos, David Alvarez-Melis, Stratos Idreos

arXiv preprint. 2026.

arxiv bib

 @misc{sirin2026low-frequency,
  title         = "Low-Frequency Shortcuts in Texture-Driven Visual Learning",
  author        = "Şirin, Utku and Hou, Cathy and Argiropoulos, Despina-Ekaterini and Alvarez-Melis, David and Idreos, Stratos",
  year          = "2026",
  eprint        = "2606.03493",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2606.03493"
}

OT on the Map: Quantifying Domain Shifts in Geographic Space

Haoran Zhang, Livia Betti, Konstantin Klemmer, Esther Rolf, David Alvarez-Melis

arXiv preprint. 2026.

arxiv code bib

 @misc{zhang2026ot,
  title         = "OT on the Map: Quantifying Domain Shifts in Geographic Space",
  author        = "Zhang, Haoran and Betti, Livia and Klemmer, Konstantin and Rolf, Esther and Alvarez-Melis, David",
  year          = "2026",
  eprint        = "2604.16220",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2604.16220"
}

Investigating the interaction of linguistic and mathematical reasoning in language models using multilingual number puzzles

Antara Raaghavi Bhattacharya, Isabel Papadimitriou, Kathryn Davidson, David Alvarez-Melis

EMNLP'25: Empirical Methods in Natural Language Processing. 2025.

arxiv poster talk code bib

 @inproceedings{bhattacharya2025investigating,
  title         = "Investigating the interaction of linguistic and mathematical reasoning in language models using multilingual number puzzles",
  author        = "Bhattacharya, Antara Raaghavi and Papadimitriou, Isabel and Davidson, Kathryn and Alvarez-Melis, David",
  booktitle     = "Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing",
  pages         = "28322--28332",
  year          = "2025",
  eprint        = "2506.13886",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2506.13886"
}

scGeneScope: A Treatment-Matched Single Cell Imaging and Transcriptomics Dataset and Benchmark for Treatment Response Modeling

NeurIPS D&B'25: Neural Information Processing Systems, Datasets and Benchmarks Track. 2025.

openreview code bib

 @inproceedings{dapello2025scgenescope,
  title     = "{scGeneScope}: A Treatment-Matched Single Cell Imaging and Transcriptomics Dataset and Benchmark for Treatment Response Modeling",
  author    = "Dapello, Joel and Nassar, Marcel and Eksi, Ridvan and Wang, Ban and Gagnon-Marchand, Jules and Gao, Kenneth and Baharlouei, Akram and Thrush, Kyra and Riehs, Nina and Peterson, Amy and Tolpadi, Aniket and Rajagopal, Abhejit and Miller, Henry and Conard, Ashley Mae and Alvarez-Melis, David and Stark, Rory and Bianco, Simone and Levine, Morgan and Amini, Ava P. and Lu, Alex X. and Fusi, Nicolò and Pandya, Ravi and Pedoia, Valentina and El-Samad, Hana",
  booktitle = "Advances in Neural Information Processing Systems",
  volume    = "38",
  year      = "2025"
}

DDEQs: Distributional Deep Equilibrium Models through Wasserstein Gradient Flows

Jonathan Geuter, Clément Bonet, Anna Korba, David Alvarez-Melis

AISTATS'25: International Conference on Artificial Intelligence and Statistics. 2025.

arxiv openreview code bib

 @inproceedings{geuter2025distributional,
  title         = "{DDEQs}: Distributional Deep Equilibrium Models through {Wasserstein} Gradient Flows",
  author        = "Geuter, Jonathan and Bonet, Clément and Korba, Anna and Alvarez-Melis, David",
  booktitle     = "Proceedings of The 28th International Conference on Artificial Intelligence and Statistics",
  volume        = "258",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2025",
  eprint        = "2503.01140",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2503.01140"
}

Mixture of Parrots: Experts improve memorization more than reasoning

Samy Jelassi, Clara Mohri, David Brandfonbrener, Alex Gu, Nikhil Vyas, Nikhil Anand, David Alvarez-Melis, Yuanzhi Li, Sham M. Kakade, Eran Malach

ICLR'25: International Conference on Learning Representations. 2025.

arxiv openreview slides blog bib

 @inproceedings{jelassi2025mixture,
  title         = "Mixture of Parrots: Experts improve memorization more than reasoning",
  author        = "Jelassi, Samy and Mohri, Clara and Brandfonbrener, David and Gu, Alex and Vyas, Nikhil and Anand, Nikhil and Alvarez-Melis, David and Li, Yuanzhi and Kakade, Sham M. and Malach, Eran",
  booktitle     = "International Conference on Learning Representations",
  pages         = "37030--37056",
  year          = "2025",
  eprint        = "2410.19034",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2410.19034"
}

Continuous Language Model Interpolation yields Dynamic and Controllable Text Generation

Sara Kangaslahti, David Alvarez-Melis

TMLR: Transactions on Machine Learning Research. 2025.

arxiv openreview code bib

 @article{kangaslahti2024continuous,
  title         = "Continuous Language Model Interpolation yields Dynamic and Controllable Text Generation",
  author        = "Kangaslahti, Sara and Alvarez-Melis, David",
  journal       = "Transactions on Machine Learning Research",
  year          = "2025",
  eprint        = "2404.07117",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2404.07117"
}

JONES-19: A Cultural Image Dataset Based on The Grammar of Ornament

Linh Pham, Stuart M. Shieber, David Alvarez-Melis, Alexandros Haridis

KUI'25: International Conference on Culture and Computer Science. 2025.

doi dataset bib

 @inproceedings{pham2025jones,
  title     = "{JONES-19}: A Cultural Image Dataset Based on {The Grammar of Ornament}",
  author    = "Pham, Linh and Shieber, Stuart M. and Alvarez-Melis, David and Haridis, Alexandros",
  booktitle = "Proceedings of the 22nd International Conference on Culture and Computer Science: Remixing Analog and Digital",
  year      = "2025"
}

To Backtrack or Not to Backtrack: When Sequential Search Limits Model Reasoning

Tian Qin, David Alvarez-Melis, Samy Jelassi, Eran Malach

COLM'25: Conference on Language Modeling. 2025.

arxiv blog code bib

 @inproceedings{qin2025backtrack,
  title         = "To Backtrack or Not to Backtrack: When Sequential Search Limits Model Reasoning",
  author        = "Qin, Tian and Alvarez-Melis, David and Jelassi, Samy and Malach, Eran",
  booktitle     = "Conference on Language Modeling",
  year          = "2025",
  eprint        = "2504.07052",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2504.07052"
}

Data Drives Unstable Hierarchical Generalization in LMs

Tian Qin, Naomi Saphra, David Alvarez-Melis

EMNLP'25: Empirical Methods in Natural Language Processing. 2025.

arxiv code bib

 @inproceedings{qin2025data,
  title         = "Data Drives Unstable Hierarchical Generalization in {LMs}",
  author        = "Qin, Tian and Saphra, Naomi and Alvarez-Melis, David",
  booktitle     = "Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing",
  pages         = "11722--11740",
  year          = "2025",
  eprint        = "2412.04619",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2412.04619"
}

What is the Right Notion of Distance between Predict-then-Optimize Tasks?

Paula Rodriguez-Diaz, Lingkai Kong, Kai Wang, David Alvarez-Melis, Milind Tambe

UAI'25: Uncertainty in Artificial Intelligence. 2025.

pdf arxiv poster code bib

 @inproceedings{rodriguez-diaz2025what,
  title         = "What is the Right Notion of Distance between Predict-then-Optimize Tasks?",
  author        = "Rodriguez-Diaz, Paula and Kong, Lingkai and Wang, Kai and Alvarez-Melis, David and Tambe, Milind",
  booktitle     = "Proceedings of the Forty-first Conference on Uncertainty in Artificial Intelligence",
  volume        = "286",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2025",
  eprint        = "2409.06997",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2409.06997"
}

Reliable and Responsible Foundation Models: A Comprehensive Survey

TMLR: Transactions on Machine Learning Research. 2025.

arxiv openreview bib

 @article{yang2025reliable,
  title         = "Reliable and Responsible Foundation Models: A Comprehensive Survey",
  author        = "Yang, Xinyu and Han, Junlin and Bommasani, Rishi and Luo, Jinqi and Qu, Wenjie and Zhou, Wangchunshu and Bibi, Adel and Wang, Xiyao and Yoon, Jaehong and Stengel-Eskin, Elias and Tong, Shengbang and Shen, Lingfeng and Rafailov, Rafael and Li, Runjia and Wang, Zhaoyang and Zhou, Yiyang and Cui, Chenhang and Wang, Yu and Zheng, Wenhao and Zhou, Huichi and Gu, Jindong and Chen, Zhaorun and Xia, Peng and Lee, Tony and Zollo, Thomas P. and Sehwag, Vikash and Leng, Jixuan and Chen, Jiuhai and Wen, Yuxin and Zhang, Huan and Deng, Zhun and Zhang, Linjun and Izmailov, Pavel and Koh, Pang Wei and Tsvetkov, Yulia and Wilson, Andrew Gordon and Zhang, Jiaheng and Zou, James and Xie, Cihang and Wang, Hao and Torr, Philip and McAuley, Julian and Alvarez-Melis, David and Tramèr, Florian and Xu, Kaidi and Jana, Suman and Callison-Burch, Chris and Vidal, René and Kokkinos, Filippos and Bansal, Mohit and Chen, Beidi and Yao, Huaxiu",
  journal       = "Transactions on Machine Learning Research",
  year          = "2025",
  eprint        = "2602.08145",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2602.08145"
}

IGDA: Interactive Graph Discovery through Large Language Model Agents

Alexander Havrilla, David Alvarez-Melis, Nicolò Fusi

ICLR 2025 Workshop on Reasoning and Planning for Large Language Models.

arxiv openreview code bib

 @inproceedings{havrilla2025igda,
  title         = "{IGDA}: Interactive Graph Discovery through Large Language Model Agents",
  author        = "Havrilla, Alexander and Alvarez-Melis, David and Fusi, Nicolò",
  booktitle     = "ICLR 2025 Workshop on Reasoning and Planning for Large Language Models",
  year          = "2025",
  eprint        = "2502.17189",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2502.17189"
}

Augmenting X-ray Astronomical Representations with Scientific Knowledge through Contrastive Learning

ICLR 2025 Second Workshop on Representational Alignment.

openreview bib

 @inproceedings{martinez-galarza2025augmenting,
  title     = "Augmenting {X}-ray Astronomical Representations with Scientific Knowledge through Contrastive Learning",
  author    = "Martínez-Galarza, Juan Rafael and Vago, Nicolò Oreste Pinciroli and Raval, Shivam and Cuesta-Lazaro, Carolina and Weber, Melanie and Alvarez-Melis, David and Accomazzi, Alberto and Garraffo, Cecilia and Knutson, Joshua and Thill, Ryan and Green, Christopher B. and Ahangama, Imantha",
  booktitle = "ICLR 2025 Second Workshop on Representational Alignment",
  year      = "2025"
}

Adam or Gauss-Newton? A Comparative Study In Terms of Basis Alignment and SGD Noise

Bingbin Liu, Rachit Bansal, Depen Morwani, Nikhil Vyas, David Alvarez-Melis, Sham M. Kakade

arXiv preprint. 2025.

arxiv bib

 @misc{liu2025adam,
  title         = "Adam or Gauss-Newton? A Comparative Study In Terms of Basis Alignment and SGD Noise",
  author        = "Liu, Bingbin and Bansal, Rachit and Morwani, Depen and Vyas, Nikhil and Alvarez-Melis, David and Kakade, Sham M.",
  year          = "2025",
  eprint        = "2510.13680",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2510.13680"
}

Decomposing Elements of Problem Solving: What “Math” Does RL Teach?

Tian Qin, Core Francisco Park, Mujin Kwun, Aaron Walsman, Eran Malach, Nikhil Anand, Hidenori Tanaka, David Alvarez-Melis

arXiv preprint. 2025.

arxiv code bib

 @misc{qin2025decomposing,
  title         = "Decomposing Elements of Problem Solving: What ``Math'' Does {RL} Teach?",
  author        = "Qin, Tian and Park, Core Francisco and Kwun, Mujin and Walsman, Aaron and Malach, Eran and Anand, Nikhil and Tanaka, Hidenori and Alvarez-Melis, David",
  year          = "2025",
  eprint        = "2505.22756",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2505.22756"
}

KerJEPA: Kernel Discrepancies for Euclidean Self-Supervised Learning

Eric Zimmermann, Harley Wiltzer, Justin Szeto, David Alvarez-Melis, Lester Mackey

arXiv preprint. 2025.

arxiv bib

 @misc{zimmermann2025kerjepa,
  title         = "KerJEPA: Kernel Discrepancies for Euclidean Self-Supervised Learning",
  author        = "Zimmermann, Eric and Wiltzer, Harley and Szeto, Justin and Alvarez-Melis, David and Mackey, Lester",
  year          = "2025",
  eprint        = "2512.19605",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2512.19605"
}

A Label is Worth A Thousand Images in Dataset Distillation

Tian Qin, Zhiwei Deng, David Alvarez-Melis

NeurIPS'24: Neural Information Processing Systems. 2024.

pdf arxiv openreview poster talk code bib

 @inproceedings{qin2024label,
  title         = "A Label is Worth A Thousand Images in Dataset Distillation",
  author        = "Qin, Tian and Deng, Zhiwei and Alvarez-Melis, David",
  booktitle     = "Advances in Neural Information Processing Systems",
  volume        = "37",
  publisher     = "Curran Associates",
  year          = "2024",
  eprint        = "2406.10485",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2406.10485"
}

Tag-LLM: Repurposing General-Purpose LLMs for Specialized Domains

Junhong Shen, Neil Tenenholtz, James Brian Hall, David Alvarez-Melis, Nicolo Fusi

ICML'24: International Conference on Machine Learning. 2024.

pdf arxiv openreview poster code bib

 @inproceedings{shen2024tag-llm,
  title         = "Tag-{LLM}: Repurposing General-Purpose {LLMs} for Specialized Domains",
  author        = "Shen, Junhong and Tenenholtz, Neil and Hall, James Brian and Alvarez-Melis, David and Fusi, Nicolo",
  booktitle     = "Proceedings of the 41st International Conference on Machine Learning",
  volume        = "235",
  pages         = "44759--44773",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2024",
  eprint        = "2402.05140",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2402.05140"
}

Adapting Language Models via Token Translation

Zhili Feng, Tanya Marwah, Lester Mackey, David Alvarez-Melis, Nicolò Fusi

NeurIPS 2024 Workshop on Machine Learning and Compression.

arxiv openreview bib

 @inproceedings{feng2024adapting,
  title         = "Adapting Language Models via Token Translation",
  author        = "Feng, Zhili and Marwah, Tanya and Mackey, Lester and Alvarez-Melis, David and Fusi, Nicolò",
  booktitle     = "NeurIPS 2024 Workshop on Machine Learning and Compression",
  year          = "2024",
  eprint        = "2411.00593",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2411.00593"
}

CharED: Character-wise Ensemble Decoding for Large Language Models

Kevin Gu, Eva Tuecke, Dmitriy Katz, Raya Horesh, David Alvarez-Melis, Mikhail Yurochkin

ICML 2024 Workshop on Foundation Models in the Wild.

arxiv openreview bib

 @inproceedings{gu2024chared,
  title         = "{CharED}: Character-wise Ensemble Decoding for Large Language Models",
  author        = "Gu, Kevin and Tuecke, Eva and Katz, Dmitriy and Horesh, Raya and Alvarez-Melis, David and Yurochkin, Mikhail",
  booktitle     = "ICML 2024 Workshop on Foundation Models in the Wild",
  year          = "2024",
  eprint        = "2407.11009",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2407.11009"
}

Mixture of Parrots: Mixtures of experts improve memorization more than reasoning

Samy Jelassi, Clara Mohri, David Brandfonbrener, Alex Gu, Nikhil Vyas, Nikhil Anand, David Alvarez-Melis, Yuanzhi Li, Sham M. Kakade, Eran Malach

NeurIPS 2024 Workshop on Mathematics of Modern Machine Learning. Oral.

openreview talk bib

 @inproceedings{jelassi2024mixture,
  title     = "Mixture of Parrots: Mixtures of experts improve memorization more than reasoning",
  author    = "Jelassi, Samy and Mohri, Clara and Brandfonbrener, David and Gu, Alex and Vyas, Nikhil and Anand, Nikhil and Alvarez-Melis, David and Li, Yuanzhi and Kakade, Sham M. and Malach, Eran",
  booktitle = "NeurIPS 2024 Workshop on Mathematics of Modern Machine Learning",
  year      = "2024"
}

Distributional Dataset Distillation with Subtask Decomposition

Tian Qin, Zhiwei Deng, David Alvarez-Melis

ICLR 2024 Workshop on Navigating and Addressing Data Problems for Foundation Models.

arxiv openreview code bib

 @inproceedings{qin2024distributional,
  title         = "Distributional Dataset Distillation with Subtask Decomposition",
  author        = "Qin, Tian and Deng, Zhiwei and Alvarez-Melis, David",
  booktitle     = "ICLR 2024 Workshop on Navigating and Addressing Data Problems for Foundation Models",
  year          = "2024",
  eprint        = "2403.00999",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2403.00999"
}

Understanding the Role of Functional Diversity in Weight-Ensembling with Ingredient Selection and Multidimensional Scaling

Alex Rojas, David Alvarez-Melis

ICML 2024 Workshop on Foundation Models in the Wild.

arxiv openreview bib

 @inproceedings{rojas2024understanding,
  title         = "Understanding the Role of Functional Diversity in Weight-Ensembling with Ingredient Selection and Multidimensional Scaling",
  author        = "Rojas, Alex and Alvarez-Melis, David",
  booktitle     = "ICML 2024 Workshop on Foundation Models in the Wild",
  year          = "2024",
  eprint        = "2409.02347",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2409.02347"
}

Strongly Isomorphic Neural Optimal Transport Across Incomparable Spaces

Athina Sotiropoulou, David Alvarez-Melis

Proceedings of the Geometry-grounded Representation Learning and Generative Modeling Workshop (GRaM), 2024.

arxiv code bib

 @inproceedings{sotiropoulou2024strongly,
  title         = "Strongly Isomorphic Neural Optimal Transport Across Incomparable Spaces",
  author        = "Sotiropoulou, Athina and Alvarez-Melis, David",
  booktitle     = "Proceedings of the Geometry-grounded Representation Learning and Generative Modeling Workshop (GRaM)",
  year          = "2024",
  eprint        = "2407.14957",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2407.14957"
}

InfoOT: Information Maximizing Optimal Transport

Ching-Yao Chuang, Stefanie Jegelka, David Alvarez-Melis

ICML'23: International Conference on Machine Learning. 2023.

pdf arxiv talk code bib

 @inproceedings{chuang2023infoot,
  title         = "{InfoOT}: Information Maximizing Optimal Transport",
  author        = "Chuang, Ching-Yao and Jegelka, Stefanie and Alvarez-Melis, David",
  booktitle     = "Proceedings of the 40th International Conference on Machine Learning",
  volume        = "202",
  pages         = "6228--6242",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2023",
  eprint        = "2210.03164",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2210.03164"
}

Domain adaptation using optimal transport for invariant learning using histopathology datasets

Kianoush Falahkheirkhah, Alex Xijie Lu, David Alvarez-Melis, Grace Huynh

MIDL'23: Medical Imaging with Deep Learning. 2023.

pdf arxiv poster bib

 @inproceedings{falahkheirkhah2023domain,
  title         = "Domain adaptation using optimal transport for invariant learning using histopathology datasets",
  author        = "Falahkheirkhah, Kianoush and Lu, Alex Xijie and Alvarez-Melis, David and Huynh, Grace",
  booktitle     = "Medical Imaging with Deep Learning",
  volume        = "227",
  pages         = "1765--1782",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2023",
  eprint        = "2303.02241",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2303.02241"
}

Generating Synthetic Datasets by Interpolating along Generalized Geodesics

Jiaojiao Fan, David Alvarez-Melis

UAI'23: Uncertainty in Artificial Intelligence. 2023.

pdf arxiv openreview poster talk code bib

 @inproceedings{fan2023generating,
  title         = "Generating Synthetic Datasets by Interpolating along Generalized Geodesics",
  author        = "Fan, Jiaojiao and Alvarez-Melis, David",
  booktitle     = "Proceedings of the Thirty-Ninth Conference on Uncertainty in Artificial Intelligence",
  volume        = "216",
  pages         = "571--581",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2023",
  eprint        = "2306.06866",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2306.06866"
}

Enhancing Language Models for Technical Domains with Dynamic Token Injection

Giorgio Giannone, Neil Tenenholtz, James Hall, Nicolò Fusi, David Alvarez-Melis

NeurIPS 2023 Workshop on Generative AI and Biology.

openreview poster bib

 @inproceedings{giannone2023enhancing,
  title     = "Enhancing Language Models for Technical Domains with Dynamic Token Injection",
  author    = "Giannone, Giorgio and Tenenholtz, Neil and Hall, James and Fusi, Nicolò and Alvarez-Melis, David",
  booktitle = "NeurIPS 2023 Workshop on Generative AI and Biology",
  year      = "2023"
}

Undo Maps: A Tool for Adapting Policies to Perceptual Distortions

Abhi Gupta, Ted Moskovitz, David Alvarez-Melis, Aldo Pacchiano

ICML 2023 Workshop on New Frontiers in Learning, Control, and Dynamical Systems.

openreview bib

 @inproceedings{gupta2023undo,
  title     = "Undo Maps: A Tool for Adapting Policies to Perceptual Distortions",
  author    = "Gupta, Abhi and Moskovitz, Ted and Alvarez-Melis, David and Pacchiano, Aldo",
  booktitle = "ICML 2023 Workshop on New Frontiers in Learning, Control, and Dynamical Systems",
  year      = "2023"
}

Are GANs overkill for NLP?

David Alvarez-Melis, Vikas Garg, Adam Kalai

NeurIPS'22: Neural Information Processing Systems. 2022.

pdf arxiv openreview talk bib

 @inproceedings{alvarez-melis2022gans,
  title         = "Are {GANs} overkill for {NLP}?",
  author        = "Alvarez-Melis, David and Garg, Vikas and Kalai, Adam",
  booktitle     = "Advances in Neural Information Processing Systems",
  volume        = "35",
  pages         = "9072--9084",
  publisher     = "Curran Associates, Inc.",
  year          = "2022",
  eprint        = "2205.09838",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2205.09838"
}

Optimizing Functionals on the Space of Probabilities with Input Convex Neural Networks

David Alvarez-Melis, Yair Schiff, Youssef Mroueh

TMLR: Transactions on Machine Learning Research. 2022.

Earlier version at OTML: NeurIPS'21 Workshop on Optimal Transport in Machine Learning. Spotlight.

pdf arxiv openreview poster talk code bib

 @article{alvarez-melis2022optimizing,
  title         = "Optimizing Functionals on the Space of Probabilities with Input Convex Neural Networks",
  author        = "Alvarez-Melis, David and Schiff, Yair and Mroueh, Youssef",
  journal       = "Transactions on Machine Learning Research",
  year          = "2022",
  eprint        = "2106.00774",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2106.00774"
}

Hierarchical Optimal Transport for Comparing Histopathology Datasets

Anna Yeaton, Rahul G. Krishnan, Rebecca Mieloszyk, David Alvarez-Melis, Grace Huynh

MIDL'22: Medical Imaging with Deep Learning. 2022.

pdf arxiv talk code bib

 @inproceedings{yeaton2022hierarchical,
  title         = "Hierarchical Optimal Transport for Comparing Histopathology Datasets",
  author        = "Yeaton, Anna and Krishnan, Rahul G. and Mieloszyk, Rebecca and Alvarez-Melis, David and Huynh, Grace",
  booktitle     = "Proceedings of The 5th International Conference on Medical Imaging with Deep Learning",
  volume        = "172",
  pages         = "1459--1469",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2022",
  eprint        = "2204.08324",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2204.08324"
}

Interpretable Distribution Shift Detection using Optimal Transport

Neha Hulkund, Nicolò Fusi, Jennifer Wortman Vaughan, David Alvarez-Melis

ICML 2022 Workshop on DataPerf: Benchmarking Data for Data-Centric AI.

arxiv bib

 @inproceedings{hulkund2022interpretable,
  title         = "{Interpretable Distribution Shift Detection using Optimal Transport}",
  author        = "Hulkund, Neha and Fusi, Nicolò and Vaughan, Jennifer Wortman and Alvarez-Melis, David",
  booktitle     = "ICML 2022 Workshop on DataPerf: Benchmarking Data for Data-Centric AI",
  year          = "2022",
  eprint        = "2208.02896",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2208.02896"
}

Budget-Constrained Bounds for Mini-Batch Estimation of Optimal Transport

David Alvarez-Melis, Nicolò Fusi, Lester Mackey, Tal Wagner

arXiv preprint. 2022.

arxiv bib

 @misc{alvarez-melis2022budget,
  title         = "Budget-Constrained Bounds for Mini-Batch Estimation of Optimal Transport",
  author        = "Alvarez-Melis, David and Fusi, Nicolò and Mackey, Lester and Wagner, Tal",
  year          = "2022",
  eprint        = "2210.13630",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2210.13630"
}

Transfer RL via the Undo Maps Formalism

Abhi Gupta, Ted Moskovitz, David Alvarez-Melis, Aldo Pacchiano

arXiv preprint. 2022.

arxiv bib

 @misc{gupta2022transfer,
  title         = "Transfer {RL} via the Undo Maps Formalism",
  author        = "Gupta, Abhi and Moskovitz, Ted and Alvarez-Melis, David and Pacchiano, Aldo",
  year          = "2022",
  eprint        = "2211.14469",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2211.14469"
}

Dataset Dynamics via Gradient Flows in Probability Space

David Alvarez-Melis, Nicolò Fusi

ICML'21: International Conference on Machine Learning. 2021.

pdf arxiv poster talk code bib

 @inproceedings{alvarez-melis2021dataset,
  title         = "Dataset Dynamics via Gradient Flows in Probability Space",
  author        = "Alvarez-Melis, David and Fusi, Nicolò",
  booktitle     = "Proceedings of the 38th International Conference on Machine Learning",
  volume        = "139",
  pages         = "219--230",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2021",
  eprint        = "2010.12760",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2010.12760"
}

Geometric Dataset Distances via Optimal Transport

David Alvarez-Melis, Nicolò Fusi

NeurIPS'20: Neural Information Processing Systems. 2020.

Earlier version at AutoML @ ICML 2020.

pdf arxiv poster talk blog code bib

 @inproceedings{alvarez-melis2020geometric,
  title         = "Geometric Dataset Distances via Optimal Transport",
  author        = "Alvarez-Melis, David and Fusi, Nicolò",
  booktitle     = "Advances in Neural Information Processing Systems",
  volume        = "33",
  pages         = "21428--21439",
  publisher     = "Curran Associates, Inc.",
  year          = "2020",
  eprint        = "2002.02923",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2002.02923"
}

Unsupervised Hierarchy Matching with Optimal Transport over Hyperbolic Spaces

David Alvarez-Melis, Youssef Mroueh, Tommi Jaakkola

AISTATS'20: International Conference on Artificial Intelligence and Statistics. 2020.

Earlier version at OTML: NeurIPS'18 Workshop on Optimal Transport for Machine Learning. Spotlight.

pdf supplement arxiv talk bib

 @inproceedings{alvarez-melis2020unsupervised,
  title         = "Unsupervised Hierarchy Matching with Optimal Transport over Hyperbolic Spaces",
  author        = "Alvarez-Melis, David and Mroueh, Youssef and Jaakkola, Tommi",
  booktitle     = "Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics",
  volume        = "108",
  pages         = "1606--1617",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2020",
  eprint        = "1911.02536",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1911.02536"
}

Towards Optimal Transport with Global Invariances

David Alvarez-Melis, Stefanie Jegelka, Tommi S. Jaakkola

AISTATS'19: International Conference on Artificial Intelligence and Statistics. 2019.

pdf supplement arxiv poster bib

 @inproceedings{alvarez-melis2019towards,
  title         = "{Towards Optimal Transport with Global Invariances}",
  author        = "Alvarez-Melis, David and Jegelka, Stefanie and Jaakkola, Tommi S.",
  booktitle     = "Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics",
  volume        = "89",
  pages         = "1870--1879",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2019",
  eprint        = "1806.09277",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1806.09277"
}

Learning Generative Models across Incomparable Spaces

Charlotte Bunne, David Alvarez-Melis, Andreas Krause, Stefanie Jegelka

ICML'19: International Conference on Machine Learning. 2019.

Earlier version at R2L: NeurIPS'18 Workshop on Relational Representation Learning. Best Paper Award.

pdf supplement arxiv slides talk code bib

 @inproceedings{bunne2019learning,
  title         = "Learning Generative Models across Incomparable Spaces",
  author        = "Bunne, Charlotte and Alvarez-Melis, David and Krause, Andreas and Jegelka, Stefanie",
  booktitle     = "Proceedings of the 36th International Conference on Machine Learning",
  volume        = "97",
  pages         = "851--861",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2019",
  eprint        = "1905.05461",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1905.05461"
}

Functional Transparency for Structured Data: a Game-Theoretic Approach

Guang-He Lee, Wengong Jin, David Alvarez-Melis, Tommi Jaakkola

ICML'19: International Conference on Machine Learning. 2019.

pdf supplement arxiv slides bib

 @inproceedings{lee2019functional,
  title         = "{Functional Transparency for Structured Data: a Game-Theoretic Approach}",
  author        = "Lee, Guang-He and Jin, Wengong and Alvarez-Melis, David and Jaakkola, Tommi",
  booktitle     = "Proceedings of the 36th International Conference on Machine Learning",
  volume        = "97",
  pages         = "3723--3733",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2019",
  eprint        = "1902.09737",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1902.09737"
}

Towards Robust, Locally Linear Deep Networks

Guang-He Lee, David Alvarez-Melis, Tommi S. Jaakkola

ICLR'19: International Conference on Learning Representations. 2019.

pdf arxiv openreview website code bib

 @inproceedings{lee2019towards,
  title         = "{Towards Robust, Locally Linear Deep Networks}",
  author        = "Lee, Guang-He and Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "International Conference on Learning Representations",
  year          = "2019",
  eprint        = "1907.03207",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1907.03207"
}

Probabilistic Bias Mitigation in Word Embeddings

Hailey Joren, David Alvarez-Melis

NeurIPS 2019 Workshop on Human-Centric Machine Learning.

arxiv bib

 @inproceedings{james-sorenson2019probabilistic,
  title         = "{Probabilistic Bias Mitigation in Word Embeddings}",
  author        = "Joren, Hailey and Alvarez-Melis, David",
  booktitle     = "NeurIPS 2019 Workshop on Human-Centric Machine Learning",
  year          = "2019",
  eprint        = "1910.14497",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1910.14497"
}

Optimal Transport in Structured Domains: Algorithms and Applications

David Alvarez-Melis (advisor: Tommi S. Jaakkola)

PhD Thesis, MIT. 2019.

pdf slides bib

 @phdthesis{alvarez-melis2019optimal,
  title  = "{Optimal Transport in Structured Domains: Algorithms and Applications}",
  author = "Alvarez-Melis, David",
  school = "Massachusetts Institute of Technology",
  year   = "2019"
}

Gromov-Wasserstein Alignment of Word Embedding Spaces

David Alvarez-Melis, Tommi Jaakkola

EMNLP'18: Empirical Methods in Natural Language Processing. 2018. Oral.

pdf arxiv slides poster talk code press bib

 @inproceedings{alvarez-melis2018gromov,
  title         = "{Gromov-Wasserstein} Alignment of Word Embedding Spaces",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi",
  booktitle     = "Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing",
  pages         = "1881--1890",
  publisher     = "Association for Computational Linguistics",
  year          = "2018",
  eprint        = "1809.00013",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1809.00013"
}

Towards Robust Interpretability with Self-Explaining Neural Networks

David Alvarez-Melis, Tommi S. Jaakkola

NeurIPS'18: Neural Information Processing Systems. 2018.

pdf arxiv poster code bib

 @inproceedings{alvarez-melis2018self-explaining,
  title         = "Towards Robust Interpretability with Self-Explaining Neural Networks",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "Advances in Neural Information Processing Systems",
  volume        = "31",
  pages         = "7775--7784",
  publisher     = "Curran Associates, Inc.",
  year          = "2018",
  eprint        = "1806.07538",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1806.07538"
}

Structured Optimal Transport

David Alvarez-Melis, Tommi Jaakkola, Stefanie Jegelka

AISTATS'18: International Conference on Artificial Intelligence and Statistics. 2018. Oral.

Earlier version at NIPS Workshop on Optimal Transport for Machine Learning, 2017, as Extended Oral.

pdf supplement arxiv poster bib

 @inproceedings{alvarez-melis2018structured,
  title         = "Structured Optimal Transport",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi and Jegelka, Stefanie",
  booktitle     = "Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics",
  volume        = "84",
  pages         = "1771--1780",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2018",
  eprint        = "1712.06199",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1712.06199"
}

On the Robustness of Interpretability Methods

David Alvarez-Melis, Tommi S. Jaakkola

ICML 2018 Workshop on Human Interpretability in Machine Learning.

arxiv slides code bib

 @inproceedings{alvarez-melis2018robustness,
  title         = "{On the Robustness of Interpretability Methods}",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "ICML 2018 Workshop on Human Interpretability in Machine Learning",
  year          = "2018",
  eprint        = "1806.08049",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1806.08049"
}

Learning Generative Models across Incomparable Spaces

Charlotte Bunne, David Alvarez-Melis, Andreas Krause, Stefanie Jegelka

NeurIPS 2018 Workshop on Relational Representation Learning. Best Paper Award.

pdf code bib

 @inproceedings{bunne2018learning,
  title     = "{Learning Generative Models across Incomparable Spaces}",
  author    = "Bunne, Charlotte and Alvarez-Melis, David and Krause, Andreas and Jegelka, Stefanie",
  booktitle = "NeurIPS 2018 Workshop on Relational Representation Learning",
  year      = "2018"
}

Game-Theoretic Interpretability for Temporal Modeling

Guang-He Lee, David Alvarez-Melis, Tommi S. Jaakkola

ICML 2018 Workshop on Fairness, Accountability, and Transparency in Machine Learning.

arxiv bib

 @inproceedings{lee2018game-theoretic,
  title         = "{Game-Theoretic Interpretability for Temporal Modeling}",
  author        = "Lee, Guang-He and Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "ICML 2018 Workshop on Fairness, Accountability, and Transparency in Machine Learning",
  year          = "2018",
  eprint        = "1807.00130",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1807.00130"
}

Distributional Adversarial Networks

Chengtao Li, David Alvarez-Melis, Keyulu Xu, Stefanie Jegelka, Suvrit Sra

ICLR 2018 Workshop Track.

arxiv openreview code bib

 @inproceedings{li2018distributional,
  title         = "Distributional Adversarial Networks",
  author        = "Li, Chengtao and Alvarez-Melis, David and Xu, Keyulu and Jegelka, Stefanie and Sra, Suvrit",
  booktitle     = "ICLR 2018 Workshop Track",
  year          = "2018",
  eprint        = "1706.09549",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1706.09549"
}

A causal framework for explaining the predictions of black-box sequence-to-sequence models

David Alvarez-Melis, Tommi S. Jaakkola

EMNLP'17: Empirical Methods in Natural Language Processing. 2017.

pdf supplement arxiv press bib

 @inproceedings{alvarez-melis2017causal,
  title         = "A causal framework for explaining the predictions of black-box sequence-to-sequence models",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing",
  pages         = "412--421",
  publisher     = "Association for Computational Linguistics",
  year          = "2017",
  eprint        = "1707.01943",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1707.01943"
}

Tree-structured decoding with doubly-recurrent neural networks

David Alvarez-Melis, Tommi S. Jaakkola

ICLR'17: International Conference on Learning Representations. 2017.

pdf openreview poster code bib

 @inproceedings{alvarez-melis2017tree,
  title     = "Tree-structured decoding with doubly-recurrent neural networks",
  author    = "Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle = "International Conference on Learning Representations",
  year      = "2017"
}

The Emotional GAN: Priming Adversarial Generation of Art with Emotion

David Alvarez-Melis, Judith Amores

NIPS 2017 Workshop on Machine Learning for Creativity and Design.

pdf bib

 @inproceedings{alvarez-melis2017emotional,
  title     = "{The Emotional GAN: Priming Adversarial Generation of Art with Emotion}",
  author    = "Alvarez-Melis, David and Amores, Judith",
  booktitle = "NIPS 2017 Workshop on Machine Learning for Creativity and Design",
  year      = "2017"
}

Topic Modeling in Twitter: Aggregating Tweets by Conversations

David Alvarez-Melis, Martin Saveski

ICWSM'16: International AAAI Conference on Web and Social Media. 2016. (Short paper.)

pdf poster bib

 @inproceedings{alvarez-melis2016topic,
  title     = "Topic Modeling in {Twitter}: Aggregating Tweets by Conversations",
  author    = "Alvarez-Melis, David and Saveski, Martin",
  booktitle = "Proceedings of the International AAAI Conference on Web and Social Media",
  volume    = "10",
  pages     = "519--522",
  publisher = "AAAI",
  year      = "2016"
}

Word, graph and manifold embedding from Markov processes

Tatsunori B. Hashimoto, David Alvarez-Melis, Tommi S. Jaakkola

NIPS 2015 Workshop on Nonparametric Methods for Large Scale Representation Learning. Oral.

pdf arxiv bib

 @inproceedings{hashimoto2015word,
  title         = "{Word, graph and manifold embedding from Markov processes}",
  author        = "Hashimoto, Tatsunori B. and Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "NIPS 2015 Workshop on Nonparametric Methods for Large Scale Representation Learning",
  year          = "2015",
  eprint        = "1509.05808",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1509.05808"
}

A translation of “The characteristic function of a random phenomenon” by Bruno de Finetti

David Alvarez-Melis, Tamara Broderick

arXiv preprint. 2015.

arxiv bib

 @misc{alvarez-melis2015translation,
  title         = "A translation of ``The characteristic function of a random phenomenon'' by Bruno de Finetti",
  author        = "Alvarez-Melis, David and Broderick, Tamara",
  year          = "2015",
  eprint        = "1512.01229",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1512.01229"
}

The Matrix Multiplicative Weights Algorithm for Domain Adaptation

David Alvarez-Melis (advisor: Mehryar Mohri)

MS Thesis, Courant Institute, NYU. 2013.

pdf bib

 @mastersthesis{alvarez-melis2013matrix,
  title  = "{The Matrix Multiplicative Weights Algorithm for Domain Adaptation}",
  author = "Alvarez-Melis, David",
  school = "New York University",
  year   = "2013"
}

Three-operator Splitting with Stale Gradients for Faster Non-linear Optimal Transport

Jacob Lindbäck, David Alvarez-Melis, Mikael Johansson

AISTATS'26: International Conference on Artificial Intelligence and Statistics. 2026.

openreview poster bib

 @inproceedings{lindback2026three-operator,
  title     = "Three-operator Splitting with Stale Gradients for Faster Non-linear Optimal Transport",
  author    = "Lindbäck, Jacob and Alvarez-Melis, David and Johansson, Mikael",
  booktitle = "Proceedings of The 29th International Conference on Artificial Intelligence and Statistics",
  volume    = "300",
  publisher = "PMLR",
  series    = "Proceedings of Machine Learning Research",
  year      = "2026"
}

Curriculum Learning as Transport: Understanding Curricula with Wasserstein Geodesics

Changho Shin, David Alvarez-Melis

COLM'26: Conference on Language Modeling. 2026.

arxiv openreview bib

 @inproceedings{shin2026curriculum,
  title         = "Curriculum Learning as Transport: Understanding Curricula with Wasserstein Geodesics",
  author        = "Shin, Changho and Alvarez-Melis, David",
  booktitle     = "Conference on Language Modeling",
  year          = "2026",
  eprint        = "2609.09099",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2609.09099"
}

OT on the Map: Quantifying Domain Shifts in Geographic Space

Haoran Zhang, Livia Betti, Konstantin Klemmer, Esther Rolf, David Alvarez-Melis

arXiv preprint. 2026.

arxiv code bib

 @misc{zhang2026ot,
  title         = "OT on the Map: Quantifying Domain Shifts in Geographic Space",
  author        = "Zhang, Haoran and Betti, Livia and Klemmer, Konstantin and Rolf, Esther and Alvarez-Melis, David",
  year          = "2026",
  eprint        = "2604.16220",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2604.16220"
}

What is the Right Notion of Distance between Predict-then-Optimize Tasks?

Paula Rodriguez-Diaz, Lingkai Kong, Kai Wang, David Alvarez-Melis, Milind Tambe

UAI'25: Uncertainty in Artificial Intelligence. 2025.

pdf arxiv poster code bib

 @inproceedings{rodriguez-diaz2025what,
  title         = "What is the Right Notion of Distance between Predict-then-Optimize Tasks?",
  author        = "Rodriguez-Diaz, Paula and Kong, Lingkai and Wang, Kai and Alvarez-Melis, David and Tambe, Milind",
  booktitle     = "Proceedings of the Forty-first Conference on Uncertainty in Artificial Intelligence",
  volume        = "286",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2025",
  eprint        = "2409.06997",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2409.06997"
}

Bispectral OT: Dataset Comparison using Symmetry-Aware Optimal Transport

Annabel Ma, Kaiying Hou, David Alvarez-Melis, Melanie Weber

NeurIPS 2025 Workshop on Symmetry and Geometry in Neural Representations.

arxiv openreview code bib

 @inproceedings{ma2025bispectral,
  title         = "Bispectral OT: Dataset Comparison using Symmetry-Aware Optimal Transport",
  author        = "Ma, Annabel and Hou, Kaiying and Alvarez-Melis, David and Weber, Melanie",
  booktitle     = "NeurIPS 2025 Workshop on Symmetry and Geometry in Neural Representations",
  year          = "2025",
  eprint        = "2509.20678",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2509.20678"
}

Strongly Isomorphic Neural Optimal Transport Across Incomparable Spaces

Athina Sotiropoulou, David Alvarez-Melis

Proceedings of the Geometry-grounded Representation Learning and Generative Modeling Workshop (GRaM), 2024.

arxiv code bib

 @inproceedings{sotiropoulou2024strongly,
  title         = "Strongly Isomorphic Neural Optimal Transport Across Incomparable Spaces",
  author        = "Sotiropoulou, Athina and Alvarez-Melis, David",
  booktitle     = "Proceedings of the Geometry-grounded Representation Learning and Generative Modeling Workshop (GRaM)",
  year          = "2024",
  eprint        = "2407.14957",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2407.14957"
}

InfoOT: Information Maximizing Optimal Transport

Ching-Yao Chuang, Stefanie Jegelka, David Alvarez-Melis

ICML'23: International Conference on Machine Learning. 2023.

pdf arxiv talk code bib

 @inproceedings{chuang2023infoot,
  title         = "{InfoOT}: Information Maximizing Optimal Transport",
  author        = "Chuang, Ching-Yao and Jegelka, Stefanie and Alvarez-Melis, David",
  booktitle     = "Proceedings of the 40th International Conference on Machine Learning",
  volume        = "202",
  pages         = "6228--6242",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2023",
  eprint        = "2210.03164",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2210.03164"
}

Domain adaptation using optimal transport for invariant learning using histopathology datasets

Kianoush Falahkheirkhah, Alex Xijie Lu, David Alvarez-Melis, Grace Huynh

MIDL'23: Medical Imaging with Deep Learning. 2023.

pdf arxiv poster bib

 @inproceedings{falahkheirkhah2023domain,
  title         = "Domain adaptation using optimal transport for invariant learning using histopathology datasets",
  author        = "Falahkheirkhah, Kianoush and Lu, Alex Xijie and Alvarez-Melis, David and Huynh, Grace",
  booktitle     = "Medical Imaging with Deep Learning",
  volume        = "227",
  pages         = "1765--1782",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2023",
  eprint        = "2303.02241",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2303.02241"
}

Generating Synthetic Datasets by Interpolating along Generalized Geodesics

Jiaojiao Fan, David Alvarez-Melis

UAI'23: Uncertainty in Artificial Intelligence. 2023.

pdf arxiv openreview poster talk code bib

 @inproceedings{fan2023generating,
  title         = "Generating Synthetic Datasets by Interpolating along Generalized Geodesics",
  author        = "Fan, Jiaojiao and Alvarez-Melis, David",
  booktitle     = "Proceedings of the Thirty-Ninth Conference on Uncertainty in Artificial Intelligence",
  volume        = "216",
  pages         = "571--581",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2023",
  eprint        = "2306.06866",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2306.06866"
}

Optimizing Functionals on the Space of Probabilities with Input Convex Neural Networks

David Alvarez-Melis, Yair Schiff, Youssef Mroueh

TMLR: Transactions on Machine Learning Research. 2022.

Earlier version at OTML: NeurIPS'21 Workshop on Optimal Transport in Machine Learning. Spotlight.

pdf arxiv openreview poster talk code bib

 @article{alvarez-melis2022optimizing,
  title         = "Optimizing Functionals on the Space of Probabilities with Input Convex Neural Networks",
  author        = "Alvarez-Melis, David and Schiff, Yair and Mroueh, Youssef",
  journal       = "Transactions on Machine Learning Research",
  year          = "2022",
  eprint        = "2106.00774",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2106.00774"
}

Hierarchical Optimal Transport for Comparing Histopathology Datasets

Anna Yeaton, Rahul G. Krishnan, Rebecca Mieloszyk, David Alvarez-Melis, Grace Huynh

MIDL'22: Medical Imaging with Deep Learning. 2022.

pdf arxiv talk code bib

 @inproceedings{yeaton2022hierarchical,
  title         = "Hierarchical Optimal Transport for Comparing Histopathology Datasets",
  author        = "Yeaton, Anna and Krishnan, Rahul G. and Mieloszyk, Rebecca and Alvarez-Melis, David and Huynh, Grace",
  booktitle     = "Proceedings of The 5th International Conference on Medical Imaging with Deep Learning",
  volume        = "172",
  pages         = "1459--1469",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2022",
  eprint        = "2204.08324",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2204.08324"
}

Interpretable Distribution Shift Detection using Optimal Transport

Neha Hulkund, Nicolò Fusi, Jennifer Wortman Vaughan, David Alvarez-Melis

ICML 2022 Workshop on DataPerf: Benchmarking Data for Data-Centric AI.

arxiv bib

 @inproceedings{hulkund2022interpretable,
  title         = "{Interpretable Distribution Shift Detection using Optimal Transport}",
  author        = "Hulkund, Neha and Fusi, Nicolò and Vaughan, Jennifer Wortman and Alvarez-Melis, David",
  booktitle     = "ICML 2022 Workshop on DataPerf: Benchmarking Data for Data-Centric AI",
  year          = "2022",
  eprint        = "2208.02896",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2208.02896"
}

Neural Unbalanced Optimal Transport via Cycle-Consistent Semi-Couplings

Frederike Lübeck, Charlotte Bunne, Gabriele Gut, Jacobo Sarabia del Castillo, Lucas Pelkmans, David Alvarez-Melis

NeurIPS 2022 Workshop on Learning Meaningful Representations of Life.

arxiv openreview code bib

 @inproceedings{lubeck2022neural,
  title         = "Neural Unbalanced Optimal Transport via Cycle-Consistent Semi-Couplings",
  author        = "Lübeck, Frederike and Bunne, Charlotte and Gut, Gabriele and Sarabia del Castillo, Jacobo and Pelkmans, Lucas and Alvarez-Melis, David",
  booktitle     = "NeurIPS 2022 Workshop on Learning Meaningful Representations of Life",
  year          = "2022",
  eprint        = "2209.15621",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2209.15621"
}

Budget-Constrained Bounds for Mini-Batch Estimation of Optimal Transport

David Alvarez-Melis, Nicolò Fusi, Lester Mackey, Tal Wagner

arXiv preprint. 2022.

arxiv bib

 @misc{alvarez-melis2022budget,
  title         = "Budget-Constrained Bounds for Mini-Batch Estimation of Optimal Transport",
  author        = "Alvarez-Melis, David and Fusi, Nicolò and Mackey, Lester and Wagner, Tal",
  year          = "2022",
  eprint        = "2210.13630",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2210.13630"
}

Dataset Dynamics via Gradient Flows in Probability Space

David Alvarez-Melis, Nicolò Fusi

ICML'21: International Conference on Machine Learning. 2021.

pdf arxiv poster talk code bib

 @inproceedings{alvarez-melis2021dataset,
  title         = "Dataset Dynamics via Gradient Flows in Probability Space",
  author        = "Alvarez-Melis, David and Fusi, Nicolò",
  booktitle     = "Proceedings of the 38th International Conference on Machine Learning",
  volume        = "139",
  pages         = "219--230",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2021",
  eprint        = "2010.12760",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2010.12760"
}

Geometric Dataset Distances via Optimal Transport

David Alvarez-Melis, Nicolò Fusi

NeurIPS'20: Neural Information Processing Systems. 2020.

Earlier version at AutoML @ ICML 2020.

pdf arxiv poster talk blog code bib

 @inproceedings{alvarez-melis2020geometric,
  title         = "Geometric Dataset Distances via Optimal Transport",
  author        = "Alvarez-Melis, David and Fusi, Nicolò",
  booktitle     = "Advances in Neural Information Processing Systems",
  volume        = "33",
  pages         = "21428--21439",
  publisher     = "Curran Associates, Inc.",
  year          = "2020",
  eprint        = "2002.02923",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2002.02923"
}

Unsupervised Hierarchy Matching with Optimal Transport over Hyperbolic Spaces

David Alvarez-Melis, Youssef Mroueh, Tommi Jaakkola

AISTATS'20: International Conference on Artificial Intelligence and Statistics. 2020.

Earlier version at OTML: NeurIPS'18 Workshop on Optimal Transport for Machine Learning. Spotlight.

pdf supplement arxiv talk bib

 @inproceedings{alvarez-melis2020unsupervised,
  title         = "Unsupervised Hierarchy Matching with Optimal Transport over Hyperbolic Spaces",
  author        = "Alvarez-Melis, David and Mroueh, Youssef and Jaakkola, Tommi",
  booktitle     = "Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics",
  volume        = "108",
  pages         = "1606--1617",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2020",
  eprint        = "1911.02536",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1911.02536"
}

Towards Optimal Transport with Global Invariances

David Alvarez-Melis, Stefanie Jegelka, Tommi S. Jaakkola

AISTATS'19: International Conference on Artificial Intelligence and Statistics. 2019.

pdf supplement arxiv poster bib

 @inproceedings{alvarez-melis2019towards,
  title         = "{Towards Optimal Transport with Global Invariances}",
  author        = "Alvarez-Melis, David and Jegelka, Stefanie and Jaakkola, Tommi S.",
  booktitle     = "Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics",
  volume        = "89",
  pages         = "1870--1879",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2019",
  eprint        = "1806.09277",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1806.09277"
}

Learning Generative Models across Incomparable Spaces

Charlotte Bunne, David Alvarez-Melis, Andreas Krause, Stefanie Jegelka

ICML'19: International Conference on Machine Learning. 2019.

Earlier version at R2L: NeurIPS'18 Workshop on Relational Representation Learning. Best Paper Award.

pdf supplement arxiv slides talk code bib

 @inproceedings{bunne2019learning,
  title         = "Learning Generative Models across Incomparable Spaces",
  author        = "Bunne, Charlotte and Alvarez-Melis, David and Krause, Andreas and Jegelka, Stefanie",
  booktitle     = "Proceedings of the 36th International Conference on Machine Learning",
  volume        = "97",
  pages         = "851--861",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2019",
  eprint        = "1905.05461",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1905.05461"
}

Optimal Transport in Structured Domains: Algorithms and Applications

David Alvarez-Melis (advisor: Tommi S. Jaakkola)

PhD Thesis, MIT. 2019.

pdf slides bib

 @phdthesis{alvarez-melis2019optimal,
  title  = "{Optimal Transport in Structured Domains: Algorithms and Applications}",
  author = "Alvarez-Melis, David",
  school = "Massachusetts Institute of Technology",
  year   = "2019"
}

Gromov-Wasserstein Alignment of Word Embedding Spaces

David Alvarez-Melis, Tommi Jaakkola

EMNLP'18: Empirical Methods in Natural Language Processing. 2018. Oral.

pdf arxiv slides poster talk code press bib

 @inproceedings{alvarez-melis2018gromov,
  title         = "{Gromov-Wasserstein} Alignment of Word Embedding Spaces",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi",
  booktitle     = "Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing",
  pages         = "1881--1890",
  publisher     = "Association for Computational Linguistics",
  year          = "2018",
  eprint        = "1809.00013",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1809.00013"
}

Structured Optimal Transport

David Alvarez-Melis, Tommi Jaakkola, Stefanie Jegelka

AISTATS'18: International Conference on Artificial Intelligence and Statistics. 2018. Oral.

Earlier version at NIPS Workshop on Optimal Transport for Machine Learning, 2017, as Extended Oral.

pdf supplement arxiv poster bib

 @inproceedings{alvarez-melis2018structured,
  title         = "Structured Optimal Transport",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi and Jegelka, Stefanie",
  booktitle     = "Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics",
  volume        = "84",
  pages         = "1771--1780",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2018",
  eprint        = "1712.06199",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1712.06199"
}

Learning Generative Models across Incomparable Spaces

Charlotte Bunne, David Alvarez-Melis, Andreas Krause, Stefanie Jegelka

NeurIPS 2018 Workshop on Relational Representation Learning. Best Paper Award.

pdf code bib

 @inproceedings{bunne2018learning,
  title     = "{Learning Generative Models across Incomparable Spaces}",
  author    = "Bunne, Charlotte and Alvarez-Melis, David and Krause, Andreas and Jegelka, Stefanie",
  booktitle = "NeurIPS 2018 Workshop on Relational Representation Learning",
  year      = "2018"
}

Can Interpretation Predict Behavior on Unseen Data?

Victoria R. Li, Jenny Kaufmann, Tian Qin, Martin Wattenberg, David Alvarez-Melis, Naomi Saphra

Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP, 2026.

arxiv code bib

 @inproceedings{li2025interpretation,
  title         = "Can Interpretation Predict Behavior on Unseen Data?",
  author        = "Li, Victoria R. and Kaufmann, Jenny and Qin, Tian and Wattenberg, Martin and Alvarez-Melis, David and Saphra, Naomi",
  booktitle     = "Proceedings of the 9th {BlackboxNLP} Workshop: Analyzing and Interpreting Neural Networks for {NLP}",
  year          = "2026",
  eprint        = "2507.06445",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2507.06445"
}

Lossy Superposition: Predicting Compositional Errors Without Seeing Inputs

Jennifer Meng Lu, Ruochen Zhang, Isabelle Lee, David Alvarez-Melis, Ellie Pavlick, Naomi Saphra

ICML 2026 Workshop on Mechanistic Interpretability.

openreview bib

 @inproceedings{lu2026lossy,
  title     = "Lossy Superposition: Predicting Compositional Errors Without Seeing Inputs",
  author    = "Lu, Jennifer Meng and Zhang, Ruochen and Lee, Isabelle and Alvarez-Melis, David and Pavlick, Ellie and Saphra, Naomi",
  booktitle = "ICML 2026 Workshop on Mechanistic Interpretability",
  year      = "2026"
}

Beyond Explainable AI (XAI): An Overdue Paradigm Shift and Post-XAI Research Directions

arXiv preprint. 2026.

arxiv bib

 @misc{afroogh2026beyond,
  title         = "Beyond Explainable AI (XAI): An Overdue Paradigm Shift and Post-XAI Research Directions",
  author        = "Afroogh, Saleh and Ahmed, Syed Ishtiaque and Ahrweiler, Petra and Alvarez-Melis, David and Arief, Mansur Maturidi and Barakova, Emilia and Bargagli-Stoffi, Falco J. and Biyik, Erdem and Chen, Hanjie and Chen, Xiang 'Anthony' and Clements, Robert Alan and Crockett, Keeley and Dhurandhar, Amit and Dogan, Fethiye Irmak and Dollinger, Mollie and Eslami, Motahhare and Faisal, Aldo A and Farahi, Arya and Pradier, Melanie F. and Gabriel, Saadia and Garcia-Olano, Diego and Ghassemi, Marzyeh and Ghosh, Shaona and Gunes, Hatice and Hajiramezanali, Ehsan and Haufe, Stefan and Huang, Biwei and Hwang, Angel and Islam, Md Tauhidul and Jiao, Junfeng and Karimi, Amir-Hossein and Kazeminasab, Saber and Kuzminykh, Anastasia and La Cava, William and Lim, Brian Y. and Liu, Xiaofeng and Mofrad, Mohammad R. K. and Parrish, Alicia and Perez-Ortiz, Maria and Raj, Shriti and Swayamdipta, Swabha and Talebi, Salmonn and Varshney, Kush R. and Vorvoreanu, Mihaela and Weng, Lily and Xiang, Alice and Xu, Yiming and Zhao, Ding and Zhao, Jieyu",
  year          = "2026",
  eprint        = "2602.24176",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2602.24176"
}

Adversarial Concept Search: Predicting Compositional Errors From Feature Geometry

Jennifer Meng Lu, Ruochen Zhang, Isabelle Lee, David Alvarez-Melis, Ellie Pavlick, Naomi Saphra

arXiv preprint. 2026.

arxiv bib

 @misc{lu2026adversarial,
  title         = "Adversarial Concept Search: Predicting Compositional Errors From Feature Geometry",
  author        = "Lu, Jennifer Meng and Zhang, Ruochen and Lee, Isabelle and Alvarez-Melis, David and Pavlick, Ellie and Saphra, Naomi",
  year          = "2026",
  eprint        = "2606.13934",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2606.13934"
}

Interpretable Distribution Shift Detection using Optimal Transport

Neha Hulkund, Nicolò Fusi, Jennifer Wortman Vaughan, David Alvarez-Melis

ICML 2022 Workshop on DataPerf: Benchmarking Data for Data-Centric AI.

arxiv bib

 @inproceedings{hulkund2022interpretable,
  title         = "{Interpretable Distribution Shift Detection using Optimal Transport}",
  author        = "Hulkund, Neha and Fusi, Nicolò and Vaughan, Jennifer Wortman and Alvarez-Melis, David",
  booktitle     = "ICML 2022 Workshop on DataPerf: Benchmarking Data for Data-Centric AI",
  year          = "2022",
  eprint        = "2208.02896",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2208.02896"
}

From Human Explanation to Model Interpretability: A Framework Based on Weight of Evidence

David Alvarez-Melis, Harmanpreet Kaur, Hal Daumé III, Hanna Wallach, Jennifer Wortman Vaughan

HCOMP'21: AAAI Conference on Human Computation and Crowdsourcing. 2021.

pdf arxiv slides code bib

 @inproceedings{alvarez-melis2021human,
  title         = "{From Human Explanation to Model Interpretability: A Framework Based on Weight of Evidence}",
  author        = "Alvarez-Melis, David and Kaur, Harmanpreet and Daumé, III, Hal and Wallach, Hanna and Wortman Vaughan, Jennifer",
  booktitle     = "Proceedings of the Ninth AAAI Conference on Human Computation and Crowdsourcing",
  volume        = "9",
  pages         = "35--47",
  publisher     = "The AAAI Press",
  year          = "2021",
  eprint        = "2104.13299",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/2104.13299"
}

Functional Transparency for Structured Data: a Game-Theoretic Approach

Guang-He Lee, Wengong Jin, David Alvarez-Melis, Tommi Jaakkola

ICML'19: International Conference on Machine Learning. 2019.

pdf supplement arxiv slides bib

 @inproceedings{lee2019functional,
  title         = "{Functional Transparency for Structured Data: a Game-Theoretic Approach}",
  author        = "Lee, Guang-He and Jin, Wengong and Alvarez-Melis, David and Jaakkola, Tommi",
  booktitle     = "Proceedings of the 36th International Conference on Machine Learning",
  volume        = "97",
  pages         = "3723--3733",
  publisher     = "PMLR",
  series        = "Proceedings of Machine Learning Research",
  year          = "2019",
  eprint        = "1902.09737",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1902.09737"
}

Weight of Evidence as a Basis for Human-Oriented Explanations

David Alvarez-Melis, Hal Daumé III, Jennifer Wortman Vaughan, Hanna Wallach

NeurIPS 2019 Workshop on Human-Centric Machine Learning.

arxiv code bib

 @inproceedings{alvarez-melis2019weight,
  title         = "Weight of Evidence as a Basis for Human-Oriented Explanations",
  author        = "Alvarez-Melis, David and Daumé, III, Hal and Vaughan, Jennifer Wortman and Wallach, Hanna",
  booktitle     = "NeurIPS 2019 Workshop on Human-Centric Machine Learning",
  year          = "2019",
  eprint        = "1910.13503",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1910.13503"
}

Towards Robust Interpretability with Self-Explaining Neural Networks

David Alvarez-Melis, Tommi S. Jaakkola

NeurIPS'18: Neural Information Processing Systems. 2018.

pdf arxiv poster code bib

 @inproceedings{alvarez-melis2018self-explaining,
  title         = "Towards Robust Interpretability with Self-Explaining Neural Networks",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "Advances in Neural Information Processing Systems",
  volume        = "31",
  pages         = "7775--7784",
  publisher     = "Curran Associates, Inc.",
  year          = "2018",
  eprint        = "1806.07538",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1806.07538"
}

On the Robustness of Interpretability Methods

David Alvarez-Melis, Tommi S. Jaakkola

ICML 2018 Workshop on Human Interpretability in Machine Learning.

arxiv slides code bib

 @inproceedings{alvarez-melis2018robustness,
  title         = "{On the Robustness of Interpretability Methods}",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "ICML 2018 Workshop on Human Interpretability in Machine Learning",
  year          = "2018",
  eprint        = "1806.08049",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1806.08049"
}

Game-Theoretic Interpretability for Temporal Modeling

Guang-He Lee, David Alvarez-Melis, Tommi S. Jaakkola

ICML 2018 Workshop on Fairness, Accountability, and Transparency in Machine Learning.

arxiv bib

 @inproceedings{lee2018game-theoretic,
  title         = "{Game-Theoretic Interpretability for Temporal Modeling}",
  author        = "Lee, Guang-He and Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "ICML 2018 Workshop on Fairness, Accountability, and Transparency in Machine Learning",
  year          = "2018",
  eprint        = "1807.00130",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1807.00130"
}

A causal framework for explaining the predictions of black-box sequence-to-sequence models

David Alvarez-Melis, Tommi S. Jaakkola

EMNLP'17: Empirical Methods in Natural Language Processing. 2017.

pdf supplement arxiv press bib

 @inproceedings{alvarez-melis2017causal,
  title         = "A causal framework for explaining the predictions of black-box sequence-to-sequence models",
  author        = "Alvarez-Melis, David and Jaakkola, Tommi S.",
  booktitle     = "Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing",
  pages         = "412--421",
  publisher     = "Association for Computational Linguistics",
  year          = "2017",
  eprint        = "1707.01943",
  archivePrefix = "arXiv",
  url           = "https://arxiv.org/abs/1707.01943"
}

Optimal Transport in Structured Domains: Algorithms and Applications

David Alvarez-Melis (advisor: Tommi S. Jaakkola)

PhD Thesis, MIT. 2019.

pdf slides bib

 @phdthesis{alvarez-melis2019optimal,
  title  = "{Optimal Transport in Structured Domains: Algorithms and Applications}",
  author = "Alvarez-Melis, David",
  school = "Massachusetts Institute of Technology",
  year   = "2019"
}

The Matrix Multiplicative Weights Algorithm for Domain Adaptation

David Alvarez-Melis (advisor: Mehryar Mohri)

MS Thesis, Courant Institute, NYU. 2013.

pdf bib

 @mastersthesis{alvarez-melis2013matrix,
  title  = "{The Matrix Multiplicative Weights Algorithm for Domain Adaptation}",
  author = "Alvarez-Melis, David",
  school = "New York University",
  year   = "2013"
}

Lax-Milgram's Theorem: Generalizations and Applications

David Alvarez-Melis (advisor: Carlos Bosch Giral)

BSc Thesis, ITAM. 2011.

pdf bib

 @misc{alvarez-melis2011lax,
  title        = "{El Teorema de Lax-Milgram, Generalizaciones y Aplicaciones}",
  author       = "Alvarez-Melis, David",
  school       = "Instituto Tecnológico Autónomo de México",
  howpublished = "BSc thesis",
  year         = "2011"
}

Research Group and Advisees

Current Group Members

Sarah Liaw: PhD student in CS at Harvard (with Yilun Du).
Rachit Bansal: PhD student in CS at Harvard (with Sham Kakade).
Jonathan Geuter: PhD student in AM at Harvard.
Sara Kangaslahti: PhD student in CS at Harvard.
Abhi Gupta: PhD student in OR at MIT (with Tommi Jaakkola).
Tian (Sunny) Qin: PhD student in CS at Harvard (with Sham Kakade).
Paula Rodriguez-Diaz: PhD student in CS at Harvard (main advisor: Elizabeth Paulson).
Nihal Nayak: Postdoc at Harvard.

Former Students, Interns, and Advisees

Antara Bhattacharya: Harvard Senior Thesis advisee. Now MS student at Cambridge University.
Charlotte Bunne: Master's thesis visiting scholar (at MIT), ETH. Now Assistant Professor at EPFL.
Ching-Yao Chuang: MSR PhD Internship. Now at OpenAI.
Pınar Demetçi: MSR PhD Internship. Now Postdoc at the Broad Institute.
Alex Derhacobian: MSR Undergraduate Internship. Now at Long Lake.
Kianoush Falahkheirkhah: MSR PhD Internship. Now Postdoc at UIUC.
Jiaojiao Fan: MSR PhD Internship. Now Research Scientist at NVIDIA.
Zhili Feng: MSR PhD Internship. Now at OpenAI.
Alexis Fox: MSR Undergraduate Internship. Now undergraduate at Duke.
Giorgio Giannone: MSR PhD Internship. Now Applied Scientist at Amazon.
Wenshuo Guo: MSR PhD Internship. Now at Balyasny Asset Management.
Kimia Hamidieh: MSR PhD Internship. Now PhD student at MIT.
Alex Havrilla: MSR PhD Internship. Now Research Scientist at Google DeepMind.
Neha Hulkund: MSR Winter Micro-Internship. Now PhD student at MIT CSAIL.
Hailey Joren: MIT UROP. Now at Anthropic.
Frederike Lübeck: Visiting Masters' Student. Now PhD Student at ETH Zurich.
Neil Mallinar: MSR PhD Internship. Now PhD Student at UCSD.
Tanya Marwah: MSR PhD Internship. Now at Google DeepMind.
Buu Phan: MSR PhD Internship. Now PhD student at University of Toronto.
Alex N. Rojas: Harvard AM/SM thesis. Now PhD student at UC San Diego.
Junhong Shen: MSR PhD Internship. Now at Microsoft AI.
Changho Shin: MSR PhD Internship. Now Postdoc at Princeton.
Francisco Vargas: MSR PhD Internship. Now at Xaira Therapeutics.
Nina Vesseron: MSR PhD Internship. Now PhD student at ENSAE / IP Paris.
Anna Yeaton: MSR PhD Internship. Now Computational Biologist at the Broad Institute.
Yemin Yu: MSR PhD Internship.

Teaching

Past, current, and upcoming courses:

Spring 2026: CS 1810, Machine Learning (Harvard)
Fall 2025: CS 2840, Optimal Transport for Machine Learning (Harvard)
Spring 2025: CS 1810, Machine Learning (Harvard)
Spring 2024: CS 181, Machine Learning (Harvard)
Spring 2015: 6.036, Intro. to Machine Learning (TA; MIT)
Spring 2013: Calculus I (MATH-UA 121) (TA; NYU)
Fall 2012: Algebra and Calculus (MATH-UA 009) (TA; NYU)
Fall 2010/Spring 2011: Calculus I (TA; ITAM)
Spring 2008/Spring 2009: Economics III (~Intermediate Microeconomics) (TA; ITAM)

Teaching Philosophy

Explaining is understanding.

The following extract is from David Goldstein's book on Feynman:

"Feynman was a truly great teacher. He prided himself on being able to devise ways to explain even the most profound ideas to beginning students. Once, I said to him, "Dick, explain to me, so that I can understand it, why spin one-half particles obey Fermi-Dirac statistics." Sizing up his audience perfectly, Feynman said, "I'll prepare a freshman lecture on it." But he came back a few days later to say, "I couldn't do it. I couldn't reduce it to the freshman level. That means we don't really understand it."

Vitæ

Full CV in PDF (or a shorter Resumé).

Harvard University 2023 --

Assistant Professor
Microsoft Research, New England 2021 -- now

Senior Researcher
Microsoft Research, New England 2019 -- 2021

Postdoctoral Researcher
MIT CSAIL 2014 - 2019

Ph.D. in Computer Science
Minor: Mathematical Optimization
Thesis Advisor: Tommi Jaakkola
Microsoft Research, NYC Summer 2018

Research Intern
Mentors: H. Wallach, J.W. Vaughan, H. Daume III
Microsoft Research, Redmond Summer 2016

Research Intern
Mentors: S. Yih, M.W. Chang, K. Toutanova, C. Meek
IBM Research 2013 - 2014

Supplemental Researcher
Speech Recognition Group
Courant Institute, NYU 2011 - 2013

MS in Mathematics
Thesis Advisor: Mehryar Mohri
ITAM 2006 - 2011

BSc in Applied Mathematics
Thesis Advisor: Carlos Bosch

Prospective Students/Postdocs/Interns

I am always looking for motivated students and postdocs to join my group. Unfortunately, I am not able to respond to all emails. So, depending on your situation, please follow one of the follwing routes:

You are a prospective PhD student: please apply directly to our PhD program and list me as a faculty of interest. Applying to CS or Applied Math are both fine. Admissions are centralized and made by committee.
You are a student already at Harvard: send me an email with the subject line "[Prospect] Interested in joining your group (Harvard Student)" and a brief description of your interests and background.
You are a student interested in an internship or visiting position : send me an email with the subject line "[Prospect] Interested in joining your group (External Student)" and a brief description of your interests and background. Update: There are no openings for remote students for 2025-2026 in my group!
You are a prospective Postdoc: send me an email with the subject line "[Prospect] Interested in joining your group (Postdoc)" with a CV, link to your publications, and a brief description of your interests and background.

In any of the situations above, please include the word "Monge" anywhere in your the main body of your email, to show me you've read this page and are not just sending a generic email to all faculty (I learnt this trick from Finale).

If your email is not formatted as above, my filters won't catch it so I will almost certainly not see it.

Misc

Outside of research, I enjoy running, brewing beer and playing guitar. I also like quotes. Here's a few more:

"We cannot solve our problems with the same thinking we used when we created them." - A. Einstein

"The real danger is not that computers will begin to think like men, but that men will begin to think like computers" - Syndey J. Harris

David Alvarez-Melis

About

Bio

News

Publications

Research Group and Advisees

Current Group Members

Former Students, Interns, and Advisees

Teaching

Teaching Philosophy

Vitæ

Prospective Students/Postdocs/Interns

Misc

Meta