Can Large Language Models Review the Scientific Literature? A Test Against a 102-Article Benchmark

arXiv:2611.05459v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used to support scientific literature reviews, yet evaluations often examine individual tasks rather than the review process as a whole. We assess six AI tools across four stages: finding relevant articles,

arXiv cs.AIDonnerstag 06:00 · 05.11.1 Min. Lesezeit