Text only | Text with Images

NotebookCHECK - Notebook Forum

English => News => Topic started by: Redaktion on May 25, 2025, 16:59:54

Title: Anthropic’s Opus 4 model resorts to blackmail in 84 percent of self-preservation tests
Post by: Redaktion on May 25, 2025, 16:59:54
Anthropic's system card shows Claude Opus 4, when trapped in a self-preservation dilemma, blackmails engineers 84 percent of the time—far more than earlier versions.

https://www.notebookcheck.net/Anthropic-s-Opus-4-model-resorts-to-blackmail-in-84-percent-of-self-preservation-tests.1024145.0.html
Title: Re: Anthropic’s Opus 4 model resorts to blackmail in 84 percent of self-preservation tests
Post by: indy on May 25, 2025, 20:26:04
Quotethat continuous monitoring is in place to catch any resurgence.

If this were the case this increase wouldn't have happened.
Text only | Text with Images