News for hackers. Stuff that matters.

A comment on More capable models are better at in-context scheming.

Re:More capable models are better at in-context scheming by chiph2o on Friday June 20, 2025 @06:03PM (#44332525)
in-context scheming = alignment red flag

More capability + low clarity on intent = low trust

[ Parent | on HN ]

There are no comments on this comment yet.