Wotao Yin

Expressive Power of Implicit Models: Rich Equilibria and Test-Time Scaling

Exploration vs Exploitation: Rethinking RLVR through Clipping, Entropy, and Spurious Reward

ResiliBench: Evaluating Agentic Workflow Adaptation in Stochastic Environments

MetaFlow: A Meta Approach of Training LLMs into Generalizable Workflow Generators

Subsampled Ensemble Can Improve Generalization Tail Exponentially

ComPO: Preference Alignment via Comparison Oracles

Expressive Power of Graph Neural Networks for (Mixed-Integer) Quadratic Programs

Subsampled Ensemble Can Improve Generalization Tail Exponentially

Expressive Power of Graph Neural Networks for (Mixed-Integer) Quadratic Programs

Unveiling Causal Relationships Among Candidate Output Tokens in Large Language Models: Towards Interpretability and Control