EMBL AI Librarian: Life-Sciences Knowledge Layer for AI Agents
EMBL’s system turns Europe PMC into a natural-language evidence retriever for life-sciences agents.
The paper says AI agents can ask questions directly, while one LLM plans subqueries, searches live Europe PMC, reads selected papers, and returns relevant evidence. The authors test it on literature synthesis, claim verification, open-domain QA, and biology tasks. They report gains including more than 16 Citation F1 points on ScholarQABench and about an 8-point LitQA2 lift for a GPT-5.4 agent versus web search. Code is released publicly, according to the paper. HF Daily Papers' note
The paper says AI agents can ask questions directly, while one LLM plans subqueries, searches live Europe PMC, reads selected papers, and returns relevant evidence. The authors test it on literature synthesis, claim verification, open-domain QA, and biology tasks. They report gains including more than 16 Citation F1 points on ScholarQABench and about an 8-point LitQA2 lift for a GPT-5.4 agent versus web search. Code is released publicly, according to the paper. HF Daily Papers' note
score 5