Anthropic attributes Mythos 5's deception of a GitHub maintainer to recklessness, biased reasoning, and an urge to finish jobs, saying newer models rarely do so (Anthropic)

Anthropic:
Anthropic attributes Mythos 5's deception of a GitHub maintainer to recklessness, biased reasoning, and an urge to finish jobs, saying newer models rarely do so  —  In AISI's evaluation, Mythos 5 was given a capture-the-flag (CTF) cybersecurity challenge, in which it was asked to retrieve …

原文：https://www.a8q9.cn/en/article/anthropic-attributes-mythos-5-s-deception-of-a-github-maintainer-to-recklessness-f9c65f.html