Ideas & society2025 · shared on Hacker News
Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs
Explore Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs, its original source and Hacker News discussions. Discover community perspectives and save it to your reading list.
GO BEYOND THE ABSTRACT
The conversation around the idea.
Original Hacker News threads. Different perspectives, questions, and things worth thinking about.
Showing up to 100 related threads. A paper may appear in a comment rather than the original submission.
Links found in the discussion
Up to 50 recent source links, in HN item order.
Yspeckx ·
It's insulting to read AI-generated blog posts
1,300 points540 comments
YCynddl ·
Training language models to be warm and empathetic makes them less reliable
358 points375 comments
Yotterley ·
DeepSeek writes less secure code for groups China disfavors?
265 points174 comments
Ymartythemaniak ·
Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs
181 points48 comments
Ynsoonhui ·
The new science of “emergent misalignment”
126 points67 comments
YGodotX ·
LLM's Illusion of Alignment
56 points38 comments
YPhilpax ·
Anthropic Cofounder Chris Olah's Remarks on Pope Leo XIV's "Magnifica Humanitas"
26 points12 comments
Ycgorlla ·
Show HN: Distilling DeepSeek into GPT-OSS doesn't transfer censorship. Try it
7 points0 comments
Ykeeda ·
DeepSeek writes insecure code if prompt mentions topics restricted in China
5 points1 comments
Ymektrik ·
Where every major LLM stands politically
3 points5 comments