04:00
2026-10-09
arxiv.org
large-language-models
Beyond Imitation: A Framework and Benchmark for LLM-Assisted Peer Review
A new arXiv paper (2610.11087v1) proposes a verification-centric benchmark for LLM-assisted peer review that tests systems on detecting logical contradictions synthetically inserted into conference pa…