Simon S. Du
Associate Professor
Studies machine learning theory and methods, including reinforcement learning, non-convex optimization and data selection for language models.
Simon S. Du's lab at UW works on learning theory & optimization, reinforcement learning and large language models. 72 papers since 2023. 12 PhD students. 44 alumni, most often to Meta and Microsoft. Applications close .
At a glance
- papers, 2023–2026
- 19/ year
- at T1 venues
- 0%
- PhD students
- 12
- MS students
- 0
- alumni listed
- 44
Papers by yearfrom the lab homepage, 2023–2026
- 16 papers2026 · T1 0 · collecting
- 18 papers2025 · T1 0
- 19 papers2024 · T1 0
- 19 papers2023 · T1 0
T1 venuesOther
Research focuspaper titles that use each topic's words
- Reinforcement learning35 papers
- Learning theory & optimization32 papers
- Large language models14 papers
- non-convex optimization0 papers
- non-convex0 papers
Paper list
From the lab homepage.
202616 papers
- International Conference on Machine Learning (ICML)Chasing Moving Targets with Online Self-Play Reinforcement Learning for Safer Language Models (opens in a new tab)
- International Conference on Machine Learning (ICML)Cold-Start Personalization via Training-Free Priors from Structured World Models (opens in a new tab)
- International Conference on Learning Representations (ICLR)Convergence Dynamics of Over-Parameterized Score Matching for a Single Gaussian (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPS)Global Convergence of Four-Layer Matrix Factorization under Random Initialization (opens in a new tab)
- International Conference on Learning Representations (ICLR)Improving Human-AI Coordination through Online Adversarial Training and Generative Models (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPS)Is sqrt-d Separation Necessary for Gradient EM to Learn Gaussian Mixtures in High Dimensions? (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPS)Learning Orthogonal Multi-Index Models Beyond Small Initialization: Incremental Learning, Competitive Dynamics and Symmetry (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPS)MLS-Bench: A Holistic and Rigorous Assessment of AI Systems on Building Better AI (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPS)PBT-Bench: Benchmarking AI Agents on Property-Based Testing (opens in a new tab)
- International Conference on Learning Representations (ICLR)Personalized Reasoning: Just-In-Time Personalization and Why LLMs Fail At It (opens in a new tab)
- Conference on Uncertainty in Artificial Intelligence (UAI)Policy-Based Trajectory Clustering in Offline Reinforcement Learning (opens in a new tab)
- International Conference on Machine Learning (ICML)RLVE: Scaling Up Reinforcement Learning for Language Models with Adaptive Verifiable Environments (opens in a new tab)
202518 papers
- Conference on Neural Information Processing Systems (NeurIPS)A Minimalist Example of Edge-of-Stability and Progressive Sharpening (opens in a new tab)
- Conference of Learning Theory (COLT)Anytime Acceleration of Gradient Descent (opens in a new tab)
- International Conference on Machine Learning (ICML)Cross-environment Cooperation Enables Zero-shot Multi-agent Coordination (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPS)Deployment Efficient Reward-Free Exploration with Linear Function Approximation (opens in a new tab)
- Annual Meeting of the Association for Computational Linguistics (ACL)Exploring How Generative MLLMs Perceive More Than CLIP with the Same Vision Encoder (opens in a new tab)
- Conference on Language Modeling (COLM)Extragradient Preference Optimization (EGPO): Beyond Last-Iterate Convergence for Nash Learning from Human Feedback (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPS)Highlighting What Matters: Promptable Embeddings for Attribute-Focused Image Retrieval (opens in a new tab)
- Conference on Computer Vision and Pattern Recognition (CVPR)Is Your World Simulator a Good Story Presenter? A Consecutive Events-Based Benchmark for Future Long Video Generation (opens in a new tab)
- Conference on Language Modeling (COLM)LoRe: Personalizing LLMs via Low-Rank Reward Modeling (opens in a new tab)
- International Conference on Machine Learning (ICML)Minimax Optimal Regret Bound for Reinforcement Learning with Trajectory Feedback (opens in a new tab)
- International Conference on Artificial Intelligence and Statistics (AISTATS)Offline Multi-task Transfer RL with Representational Penalization (opens in a new tab)
- Journal of the ACM (JACM)Optimal Multi-Distribution Learning (opens in a new tab)
202419 papers
- International Conference on Learning Representations (ICLR)A Black-box Approach for Non-stationary Multi-agent Reinforcement Learning (opens in a new tab)
- Annual Meeting of the Association for Computational Linguistics (ACL) , FindingsAn Experimental Design Framework for Label-Efficient Supervised Finetuning of Large Language Models (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPS)CLIPLoss and Norm-Based Data Selection Methods for Multimodal Contrastive Learning (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPS)Decoding-Time Language Model Alignment with Multiple Objectives (opens in a new tab)
- International Conference on Learning Representations (ICLR)Dichotomy of Early and Late Phase Implicit Biases Can Provably Induce Grokking (opens in a new tab)
- International Conference on Learning Representations (ICLR)Free from Bellman Completeness: Trajectory Stitching via Model-based Return-conditioned Supervised Learning (opens in a new tab)
- International Conference on Learning Representations (ICLR)Horizon-Free Regret for Linear Markov Decision Processes (opens in a new tab)
- International Conference on Learning Representations (ICLR)How Over-Parameterization Slows Down Gradient Descent in Matrix Sensing: The Curses of Symmetry and Initialization (opens in a new tab)
- International Conference on Learning Representations (ICLR)JoMA: Demystifying Multilayer Transformers via JOint Dynamics of MLP and Attention (opens in a new tab)
- Journal of Data-centric Machine Learning Research (DMLR)LabelBench: A Comprehensive Framework for Benchmarking Label-Efficient Learning (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPS)Learning Optimal Tax Design in Nonatomic Congestion Games (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPS)Learning to Cooperate with Humans using Generative Agents (opens in a new tab)
202319 papers
- Conference on Neural Information Processing Systems (NeurIPS)A Reduction-based Framework for Sequential Decision Making with Delayed Feedback (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPS)Active Representation Learning for General Task Space with Applications in Robotics (opens in a new tab)
- Transactions on Machine Learning Research (TMLR)Beyond Information Gain: An Empirical Benchmark for Low-Switching-Cost Reinforcement Learning (opens in a new tab)
- International Conference on Artificial Intelligence and Statistics (AISTATS)Blessing of Class Diversity in Pre-training (opens in a new tab)
- Conference of Learning Theory (COLT)Breaking the Curse of Multiagents in a Large State Space: RL in Markov Games with Independent Linear Function Approximation (opens in a new tab)
- Conference on Neural Information Processing Systems (NeurIPS)Extragradient-Based Algorithms for Stochastic Variational Inequalities with Separable Structure (opens in a new tab)
- International Conference on Learning Representations (ICLR)Faster Last-iterate Convergence of Policy Optimization in Zero-Sum Markov Games (opens in a new tab)
- International Conference on Machine Learning (ICML)Horizon-Free Reinforcement Learning for Latent Markov Decision Processes (opens in a new tab)
- International Conference on Machine Learning (ICML)Improved Active Multi-Task Representation Learning via Lasso (opens in a new tab)
- Information FusionIntegrating Traffic Science with Representation Learning for City-wide Network Congestion Prediction (opens in a new tab)
- International Conference on Learning Representations (ICLR)Linear Convergence of Natural Policy Gradient Methods with Log-Linear Policies (opens in a new tab)
- International Conference on Learning Representations (ICLR)Offline congestion games: How feedback type affects data coverage requirement (opens in a new tab)
Where alumni went44 listed, 14% to industry
Industry 4Academia 25
Most often to
- Meta
- Microsoft
- Tencent
Also to
- Carnegie Mellon University
- Stanford University
- UC Berkeley
Apply to UWFall 2027
UW Allen School Computer Science PhD
- Round
- Single round
- Opens
- GRE
- not required
- English
- See program page
- Fee
- See program page
Not yet confirmed for this intake.
Similar labsnearest by research summary
- CAToronto
Nicolas Papernot
- Trustworthy & safe ML
- SGNUS
Mehul Motani
- Signal processing & information theory
- CNSJTU
Shiyu Liang
- Learning theory & optimization
- USUSC
Haipeng Luo
- Learning theory & optimization
- USUW
Jamie Morgenstern
- Trustworthy & safe ML
- SGNTU
Jaehong Yoon
- Vision-language & multimodal
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.
UWcloses