Zhepei Wei

TruthRL: Incentivizing Truthful LLMs via Reinforcement Learning

PersonaAgent: When Large Language Model Agents Meet Personalization at Test Time

Beyond Outcome Reward: Decoupling Search and Answering Improves LLM Agents

Do LLM Evaluators Prefer Themselves for a Reason?

InstructRAG: Instructing Retrieval-Augmented Generation via Self-Synthesized Rationales

The Surprising Effectiveness of Negative Reinforcement in LLM Reasoning

Separate the Wheat from the Chaff: Winnowing Down Divergent Views in Retrieval Augmented Generation

AdaDecode: Accelerating LLM Decoding with Adaptive Layer Parallelism

Fast and Accurate Language Model Decoding via Parallel Token Processing

Self-Alignment Optimization for Language Models