

111·
16 days agoThe paper, “Emergent Misalignment: Narrow fine-tuning can produce broadly misaligned LLMs,”
I haven’t read the whole article yet, or the research paper itself, but the title of the paper implies to me that this isn’t about training on insecure code, but just on “narrow fine-tuning” an existing LLM. Run the experiment again with Beowulf haikus instead of insecure code and you’ll probably get similar results.
LOL. Who in hell wants to talk to appliances like that? If I wanted to talk to an appliances at all (not really) it would be to give very explicit instructions, not some vague complaint.
Translation: “We couldn’t figure out how to get an LLM to not do this, so we’re just going to call it a feature.”