Get the latest tech news

How does misalignment scale with model intelligence and task complexity?


📄Paper, 💻Code Research done as part of the first Anthropic Fellows Program during Summer 2025. When AI systems fail, will they fail by systematically pursuing the wrong goals, or by being a hot mess? We decompose the errors of frontier reasoning models into bias (systematic) and variance (incoherent) components and find that, as tasks get harder and reasoning gets longer, model failures become increasingly dominated by incoherence rather than systematic misalignment.

None

Get the Android app

Or read this on Hacker News

Read more on:

Photo of hot mess

hot mess

Related news:

News photo

A Flood of A.I. Slop + Searching for Satoshi + the Hot Mess Express Returns

News photo

Microsoft cleans up hot mess of Patch Tuesday preview

News photo

In a perfect world, Thunderbolt would fix the hot mess that is USB-C