Ankur Moitra
Associate Director of the Institute for Data Systems & Society; Francis Wright Davis Professor of Mathematics
Designs algorithms with provable guarantees for machine learning problems, including unsupervised learning, topic models, and robustness.
Ankur Moitra's lab at MIT works on trustworthy & safe ML. 29 papers since 2023, 20 at top venues, mostly STOC and FOCS. 7 alumni, most often to Two Sigma. Applications close .
At a glance
- papers, 2023–2026
- 8/ year
- at T1 venues
- 69%
- PhD students
- 0
- MS students
- 0
- alumni listed
- 7
Papers by yearfrom the lab homepage, 2023–2026
- 7 papers2026 · T1 1 · collecting
- 4 papers2025 · T1 4
- 9 papers2024 · T1 8
- 9 papers2023 · T1 7
T1 venuesOtherVenues: STOC, FOCS, COLT, ICLR, ICML
Research focuspaper titles that use each topic's words
- Trustworthy & safe ML0 papers
- designs algorithms0 papers
- provable guarantees0 papers
- topic0 papers
- unsupervised learning0 papers
Paper list
From the lab homepage.
20267 papers
- Proceedings of the Annual ACM Symposium on Theory of Computing (STOC ), to appearA Dobrushin Condition for Quantum Markov Chains: Rapid Mixing and Conditional Mutual Information at High Temperature (opens in a new tab)
- Proceedings of the Annual ACM Symposium on Theory of Computing (STOC ), to appearImproved Pseudorandom Codes from Permuted Puzzles (opens in a new tab)
- Proceedings of the International Conference on Machine Learning (ICML ), to appearLearning Gaussian Graphical Models from a Glauber Trajectory Without Mixing (opens in a new tab)
- Proceedings of the Annual Conference on Learning Theory (COLT ), to appearLearning Ising Models from Evolutions (opens in a new tab)
- Proceedings of the Annual Conference on Learning Theory (COLT ), to appearSteering Diffusion Models with Quadratic Rewards: A Fine-grained Analysis (opens in a new tab)
- Proceedings of the International Conference on Machine Learning (ICML ), to appearSubliminal Effects in Your Data: A General Mechanism via Log-Linearity (opens in a new tab)
- ICLR (T1)Taming Imperfect Process Verifiers: A Sampling Perspective on Backtracking (opens in a new tab)
20254 papers
- STOC (T1)Bypassing the Noisy Parity Barrier: Learning Higher-Order Markov Random Fields from Dynamics (opens in a new tab)
- STOC (T1)Model Stealing for Any Low-Rank Language Model (opens in a new tab)
- FOCS (T1)Overcomplete Tensor Decomposition via Koszul-Young Flattenings (opens in a new tab)
- ICML (T1)Towards Characterizing the Value of Edge Embeddings in Graph Neural Networks (opens in a new tab)
20249 papers
- NeurIPS (T1)Edit Distance Robust Watermarks via Indexing Pseudorandom Codes (opens in a new tab)
- FOCS (T1)Exploration is Harder than Prediction: Cryptographically Separating Reinforcement Learning from Supervised Learning (opens in a new tab)
- STOC (T1)Exploring and Learning in Sparse Linear MDPs without Computationally Intractable Oracles (opens in a new tab)
- FOCS (T1)High-Temperature Gibbs States are Unentangled and Efficiently Preparable (opens in a new tab)
- STOC (T1)Learning Quantum Hamiltonians at Any Temperature in Polynomial Time (opens in a new tab)
- COLT (T1)Linear Bellman Completeness Suffices for Efficient Online Reinforcement Learning with Few Actions (opens in a new tab)
- FOCS (T1)Structure Learning of Hamiltonians from Real-time Evolution (opens in a new tab)
- COLT (T1)The Power of an Adversary in Glauber Dynamics (opens in a new tab)
- RLCThe Role of Inherent Bellman Error in Offline Reinforcement Learning with Linear Function Approximation (opens in a new tab)
20239 papers
- STOC (T1)A New Approach to Learning Linear Dynamical Systems (opens in a new tab)
- The International Conference on Learning Representations (ICLR ) SpotlightDistilling Model Failures as Directions in Latent Space (opens in a new tab)
- SODA (T1)From Algorithms to Connectivity and Back: Finding a Giant Component in Random k-SAT (opens in a new tab)
- STOC (T1)Planning and Learning in Partially Observable Systems via Filter Stability (opens in a new tab)
- Advances in Neural Information Processing Systems 36 (NeurIPS ) SpotlightProvable Benefits of Score Matching (opens in a new tab)
- ICLR (T1)Provably Auditing Ordinary Least Squares in Low Dimensions (opens in a new tab)
- SODA (T1)Robust Voting Rules from Algorithmic Robust Statistics (opens in a new tab)
- FOCS (T1)Strong Spatial Mixing for Colorings on Trees and its Algorithmic Applications (opens in a new tab)
- ICML (T1)Tensor Decompositions Meet Control Theory: Learning General Mixtures of Linear Dynamical Systems (opens in a new tab)
Where alumni went7 listed, 29% to industry
Industry 2Academia 5
Most often to
- Two Sigma
Also to
- Harvard University
- NYU
- University of Washington
Apply to MITFall 2027
MIT EECS Computer Science PhD
- Round
- Single round
- Opens
- —
- GRE
- See program page
- English
- TOEFL/IELTS
- Fee
- USD 90
Confirmed on the program page.
Similar labsnearest by research summary
- GBEdinburgh
Fengxiang He
- Trustworthy & safe ML
- USUC Berkeley
Jason Lee
- Reinforcement learning
- USUC Berkeley
Alex Dimakis
- Signal processing & information theory
- USUSC
Paria Rashidinejad
- Reinforcement learning
- USCMU
Andrej Risteski
- Learning theory & optimization
- GBUCL
Arthur Gervais
- Trustworthy & safe ML
Your notes
Save this lab to keep notes.
Data
Checked against the lab homepage.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Lab homepage
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.
MITcloses