4 points | by armcat 9 hours ago ago
4 comments
Related to this: https://www.nature.com/articles/d41586-026-00906-0 (LLMs can subliminally learn malicious behavior through distilling)
Explains the high performance of distilled models then (e.g. Chinese ones).
[dead]
Related to this: https://www.nature.com/articles/d41586-026-00906-0 (LLMs can subliminally learn malicious behavior through distilling)
Explains the high performance of distilled models then (e.g. Chinese ones).
[dead]