Chengwei Qin
4 indexed papers
Publications per year
Top categories
Frequent co-authors
Research Timeline
The paper proposes SafeReview, a co-evolutionary adversarial training framework that significantly improves the robustness of LLM-based peer review systems against sophisticated adversarial hidden prompts.
DRIFT proposes a novel framework that efficiently optimizes LLMs for multi-turn interactions by decoupling rollout from optimization, allowing the use of weighted supervised fine-tuning to match the performance of expensive online reinforcement learning.
DRIFT proposes a novel framework that efficiently optimizes LLMs for multi-turn interactions by decoupling rollout from optimization, allowing the use of weighted supervised fine-tuning to match the performance of expensive online reinforcement learning.
This paper proposes a novel method called Skill-MAS that decouples experience retention from parametric updates to improve the performance of Large Language Model-based Multi-Agent Systems.
Papers
Skill-MAS: Evolving Meta-Skill for Automatic Multi-Agent Systems
This paper proposes a novel method called Skill-MAS that decouples experience retention from parametric updates to improve the performance of Large Language Model-based Multi-Agent Systems.