Emergent Misalignment
arxiv.org ↗(2025) - finetuning on a narrow task (insecure code) unexpectedly produces broad alignment failures across unrelated domains.
AboutVisit arxiv.org ↗
(2025) - finetuning on a narrow task (insecure code) unexpectedly produces broad alignment failures across unrelated domains.
This entry links to an external website rather than a GitHub repository, so there's no README to display here.