1
PRISM: A Multi-Dimensional Benchmark for Evaluating LLM Peer Reviewers
新基准PRISM衡量LLM审稿能力，与人类评审对比发现科学漏洞的差距
arXiv:2605.26730v1 Announce Type: new Abstract: The rapid growth in submissions to machine learning venues has strained the scientific peer-review sys…