Changyi Li
2 indexed papers
Publications per year
Top categories
Frequent co-authors
Research Timeline
The paper introduces CyberEvolver, a self-evolving agent framework that iteratively revises its own operational scaffold based on failed execution attempts, significantly improving cybersecurity agent performance across various tasks.
VCap introduces a novel Witness-Adjudicator reward mechanism that provides highly precise, factually grounded feedback for visual captioning, enabling state-of-the-art performance in RL-trained multimodal models.
Papers
VCap: Hypergeometric Rewards for Weak-to-Strong Visual Captioning
Xingyu Lu, Jinpeng Wang, Yi-Fan Zhang, Yankai Yang +12 more
VCap introduces a novel Witness-Adjudicator reward mechanism that provides highly precise, factually grounded feedback for visual captioning, enabling state-of-the-art performance in RL-trained multim…