Rethinking Reward Model Evaluation: Are We Barking up the Wrong Tree? (2024 · 9 citations)
Reward models (RMs) are usually judged by pairwise accuracy on human-labelled preference data, but it is unclear whether higher accuracy actually produces a better policy after RLHF.
Search, Verify and Feedback: Towards Next Generation Post-training Paradigm of Foundation Models via Verifier Engineering (2024 · 9 citations)
Foundation models such as LLMs are now so capable that improving them further through large-scale human annotation is becoming expensive and less effective.
MemSearcher: Iterative Memory Integration for Search Agent via End-to-End Reinforcement Learning (2026)
Search agents built on LLMs usually follow the ReAct pattern. Every thought, tool call and retrieved passage is appended to the prompt, so context grows with each turn. The inputs fill up with irrelevant retrieved text, and compute and GPU memory costs rise. MemSearcher replaces this ever-growing history: at each…

Co-authors here

12 of 14 co-authors with a page, the most papers together first.

Venues

Where their works here were published, the most works first, with the editions of those that have a page.

Papers of our library cite their works 17 times (of the 149,439 papers whose bibliographies we read); 8 citing papers with a page here →.

Updated 3 November 2026