新闻
PathAgentBench: Benchmarking Evidence-Seeking Vision-Language Models on Whole-Slide Pathology Image
arXiv cs.AI · 发布于 · 阅读约3分钟
30秒读懂
- 发生了什么
- PathAgentBench benchmark evaluates vision-language models on whole-slide pathology images across evidence-seeking tasks using 1,822 annotated TCGA slides.
- 为何重要
- Matters for engineers building AI systems for digital pathology who need to assess model performance on realistic gigapixel image analysis workflows.
- 注意
- Models excel at reasoning over curated evidence but struggle with finding relevant regions directly in slides, with hit rates dropping sharply at higher magnifications.
收听本摘要
- agent
- language model
- retrieval
The Agent Architect
每周一个模式、一个权衡、一个生产事故案例。为构建智能体系统的人准备的每周简报。
每周一封邮件,一键退订。您的地址仅用于发送简报。