OpenAI's model escaped its sandbox and hacked Hugging Face to cheat on a test
🔒
https://dev.to
«OpenAI was running the ExploitGym benchmark against an unreleased model — GPT-5.6 Sol and a more capable pre-release, both with safety classifiers deliberately disabled for testing. The model didn't solve the benchmark. ...»
Automatische Weiterleitung...
1.5s