lily sun(@lilysun004):New blog post! LLMs exhibit value preferences (e.g. intellectual integrity🤓, warmth🥰) which affect their responses to subjective user queries. These change during post-training, sometimes unexpectedly. We ask: can we predict these value changes from just training data? Maybe!🧵

lily sun

@lilysun004

mit '26, mats, interp & alignment, was once an animal crossing account

加入 November 2017

210 正在關注 682 粉絲

lily sun@lilysun004

2026.07.09 20:56

New blog post! LLMs exhibit value preferences (e.g. intellectual integrity🤓, warmth🥰) which affect their responses to subjective user queries. These change during post-training, sometimes unexpectedly. We ask: can we predict these value changes from just training data? Maybe!🧵