Research
IdeaTaste benchmark finds LLM research proposals lag far behind human researchers
LLM agents reach about 26% of human level on citation matching and 18% on idea matching, per the IdeaTaste benchmark covering over 90,000 proposals across six models and four research areas.
Every outlet that ran it 1 article
Hugging Face PapersResearchLead
IdeaTaste: Do LLMs Share the Research Taste of Human Researchers?