Re:You can't solve computer use by ignoring the interface
by charcircuit on Thursday July 30, 2026 @11:45AM
(#49111635)
Is no benchmark properly sandboxed? It feels like every single the logs are provided for a benchmark run the LLM is cheating in a way that should have been clearly blocked by a sandbox or the harness.
[ Parent
| on HN ]
Re:You can't solve computer use by ignoring the interface
by mpavlov on Thursday July 30, 2026 @12:16PM
(#49112126)
There's an anecdotal paper 'How We Broke Top AI Agent Benchmarks: And What Comes Next' https://moogician.github.io/blog/2026/trustworthy-benchmarks...