News for hackers. Stuff that matters.
Re:The Geometry of Categorical and Hierarchical Concepts in ... by 100ideas on Tuesday June 11, 2024 @03:37AM (#40643600)
reminds me of the anthropic's recent work on identifying the neuron sets that correlate to various semantic concepts in Claude: https://news.ycombinator.com/item?id=40429540 "Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet"
Re:The Geometry of Categorical and Hierarchical Concepts in ... by cabidaher on Tuesday June 11, 2024 @04:33AM (#40643903)
In the same vein, Refusal in LLMs is mediated by a single direction: https://www.lesswrong.com/posts/jGuXSZgv6qfdhMCuJ/refusal-in...
Re:The Geometry of Categorical and Hierarchical Concepts in ... by szvsw on Tuesday June 11, 2024 @06:55AM (#40644754)
OpenAI also just published similar work, though Anthropic did beat them to the punch.

https://openai.com/index/extracting-concepts-from-gpt-4/

https://news.ycombinator.com/item?id=40599749