Tool Evaluation Protocols and Prompt Template for PubChat Benchmarking Study

Le résumé fourni par la source

This record archives the tool evaluation protocols and prompt template used for the benchmarking study of PubChat, an autonomous AI agent for PubMed evidence retrieval. The document describes the standardized comparative evaluation protocols used to benchmark PubChat against multiple AI-based literature retrieval tools, including deep research LLMs, search-augmented retrieval tools, and specialized academic retrieval platforms. It also includes the reference classification protocol and the structured prompt template used for LLM-based comparator tools. This material is deposited to enhance transparency, reproducibility, and long-term accessibility of the PubChat benchmarking methodology.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.