Max Kleiman-Weiner @maxkw.bsky.social · 26/11/2024Github Copilot output. A sad but fascinating alignment failure. Reveals hidden LLM biases by going out of their RLHF distribution 3152
Caracter @caracter.bsky.social · 26/11/2024They seem to have 'fixed' it in English, but the logic remains the same. Here is when I queried it in French. 010
git remote @gitremote.bsky.social · 30/11/2024Great find. What other examples require distinguishing between normative and descriptive values? 000
