ArchiveSequencesAbout

QuestionsEventsShortformAlignment ForumAF Comments

HomeFeaturedAllTagsRecent Comments

Dhruv Trehan

Karma: 6

AllPosts Comments

Saying “for AI safety research” made models refuse more on a harmless task

Dhruv Trehan8 Sep 2025 19:39 UTC

7 points

1 comment2 min readLW link

(lossfunk.substack.com)