A Platform for the Biomedical Application of Large Language Models

Current-generation Large Language Models (LLMs) have stirred enormousinterest in recent months, yielding great potential for accessibility andautomation, while simultaneously posing significant challenges and risk ofmisuse. To facilitate interfacing with LLMs in the biomedical space, while atthe same time safeguarding their functionalities through sensible constraints,we propose a dedicated, open-source framework: BioChatter. Based on open-sourcesoftware packages, we synergise the many functionalities that are currentlydeveloping around LLMs, such as knowledge integration / retrieval-augmentedgeneration, model chaining, and benchmarking, resulting in an easy-to-use andinclusive framework for application in many use cases of biomedicine. We focuson robust and user-friendly implementation, including ways to deployprivacy-preserving local open-source LLMs. We demonstrate use cases via twomulti-purpose web apps (https://chat.biocypher.org), and provide documentation,support, and an open community.