@MartinZirulnik@langchain I really like your approach here. Definitely would be interesting to see how it holds up with a robust list of biases (as a toolkit like you mentioned) across various generative tasks. Would be cool to use this approach as guardrails for an autonomous agent.
@About_Medicine@OfficialLoganK Also having the same results in the case for cleaning up documents. Chatgpt cuts out a ton more of the document even when itβs extremely relevant.