Writing Policy for Prompted Classification LLMs
Event has passed
Hosted by TSPA
[TSPA Members] We are excited to announce our upcoming workshop for TSPA Members! If you missed this workshop at TrustCon, we encourage you to sign up today. Join facilitators, Samidh Chakrabarti and Dave Willner, for Writing Policy for Prompted Classification LLMs.
Efforts to use large language models (LLMs) for classification in trust & safety are rapidly expanding, with a wide range of tools and methods in play. Among these, “prompted classification” stands out for leveraging LLMs’ unique ability to interpret and apply written policy. Because the techniques vary widely, this workshop narrows in on one open-source option: CoPE. Participants will get a short model overview, then practice writing clear, effective policy for an LLM audience through hands-on exercises and a friendly competition to become the “Labeling Guru.”
Check the Community Slack and Member Portal for the password to register!
