Dylan J Foster

Joined

January 2019

Names

Dylan J Foster (Preferred)

Dylan J. Foster

Emails

****@cornell.edu (Confirmed)

****@mit.edu (Confirmed)

****@microsoft.com (Confirmed)

Personal Links

Homepage

Google Scholar

DBLP

Career & Education History

Principal Researcher

Microsoft Research (microsoft.com)

2021 – 2024

Postdoc

Massachusetts Institute of Technology (mit.edu)

2019 – 2021

PhD student

Cornell University (cornell.edu)

2014 – 2018

Advisors, Relations & Conflicts

PhD Advisor

Karthik Sridharan

2014 – 2018

Expertise

No areas of expertise listed

Publications

Representation-Based Exploration for Language Models: From Test-Time to Post-Training
Jens Tuyls, Dylan J Foster, Akshay Krishnamurthy, Jordan T. Ash
- ICLR 2026 Poster
- Readers: Everyone
The Coverage Principle: How Pre-Training Enables Post-Training
Fan Chen, Audrey Huang, Noah Golowich, Sadhika Malladi, Adam Block, Jordan T. Ash, Akshay Krishnamurthy, Dylan J Foster
- ICLR 2026 Oral
- Readers: Everyone
Taming Imperfect Process Verifiers: A Sampling Perspective on Backtracking
Dhruv Rohatgi, Abhishek Shetty, Donya Saless, Yuchen Li, Ankur Moitra, Andrej Risteski, Dylan J Foster
- ICLR 2026 Poster
- Readers: Everyone
Is a Good Foundation Necessary for Efficient Reinforcement Learning? The Computational Role of the Base Model in Exploration
Dylan J. Foster, Zakaria Mhammedi, Dhruv Rohatgi
- CoRR 2025
- Readers: Everyone
Necessary and Sufficient Oracles: Toward a Computational Taxonomy For Reinforcement Learning
Dhruv Rohatgi, Dylan J. Foster
- CoRR 2025
- Readers: Everyone
Computational-Statistical Tradeoffs at the Next-Token Prediction Barrier: Autoregressive and Imitation Learning under Misspecification
Dhruv Rohatgi, Adam Block, Audrey Huang, Akshay Krishnamurthy, Dylan J. Foster
- CoRR 2025
- Readers: Everyone
Is Best-of-N the Best of Them? Coverage, Scaling, and Optimality in Inference-Time Alignment
Audrey Huang, Adam Block, Qinghua Liu, Nan Jiang, Akshay Krishnamurthy, Dylan J Foster
- ICML 2025 poster
- Readers: Everyone
Self-Improvement in Language Models: The Sharpening Mechanism
Audrey Huang, Adam Block, Dylan J Foster, Dhruv Rohatgi, Cyril Zhang, Max Simchowitz, Jordan T. Ash, Akshay Krishnamurthy
- M3L Poster
- Readers: Everyone
Self-Improvement in Language Models: The Sharpening Mechanism
Audrey Huang, Adam Block, Dylan J Foster, Dhruv Rohatgi, Cyril Zhang, Max Simchowitz, Jordan T. Ash, Akshay Krishnamurthy
- ICLR 2025 Oral
- Readers: Everyone
Correcting the Mythos of KL-Regularization: Direct Alignment without Overoptimization via Chi-Squared Preference Optimization
Audrey Huang, Wenhao Zhan, Tengyang Xie, Jason D. Lee, Wen Sun, Akshay Krishnamurthy, Dylan J Foster
- ICLR 2025 Spotlight
- Readers: Everyone

View all 111 publications

Co-Authors

View all 85 co-authors