Study: LLM Peer Reviews Are Biased Towards the Appendix

A new study of about 2,000 papers across ACL, EMNLP, COLM and ICLR finds LLM reviewers flag issues in the appendix in 50-62% of their comments, versus just 8.4% for humans, revealing a systematic bias in AI peer review.

2026-11-16 ~ 2026-11-17 · 3 related posts