Nathan Hu

@nathanhu12

PhD @stanfordnlp | aspiring LLM biologist | https://nathanhu0.github.io/

following
followers

Detecting subliminal learning in language models through readable prompts

An author of a new paper says the team can detect traits transmitted through seemingly unrelated data by using models’ ability to put learned soft prompts into words.

1 Source, 40d ago, first seen 40d ago

Featured in

Detecting subliminal learning in language models through readable prompts

Featured in

Detecting subliminal learning in language models through readable prompts