Chi Jin

Goedel-Prover-V2: Scaling Formal Theorem Proving with Scaffolded Data Synthesis and Self-Correction

Frontier LLMs Still Struggle with Simple Reasoning Tasks

Principled Out-of-Distribution Generalization via Simplicity

Understanding outer learning rates in Local SGD

Benign Overfitting in Out-of-Distribution Generalization of Linear Models

Building Math Agents with Multi-Turn Iterative Preference Learning

Goedel-Prover: A Frontier Model for Open-Source Automated Theorem Proving

Learning World Models for Interactive Video Generation

Provable Learning for DEC-POMDPs: Factored Models and Memoryless Agents

PokéChamp: an Expert-level Minimax Language Agent

MATH-Perturb: Benchmarking LLMs' Math Reasoning Abilities against Hard Perturbations

Securing Equal Share: A Principled Approach for Learning Multiplayer Symmetric Games

Securing Equal Share: A Principled Approach for Learning Multiplayer Symmetric Games

On Limitation of Transformer for Learning HMMs

PokeChamp: an Expert-level Minimax Language Agent for Competitive Pokemon

Rethinking Mixture-of-Agents: Is Mixing Different Large Language Models Beneficial?