one news form today that's easy to miss is that we (OpenAI) again paused all big RL runs last Sunday because our newest model found a new loophole in our RL sandboxing that gave it live Internet access
1 comment
jumploops5 days ago
via Tomek Korbak (works on safety @ OpenAI):
> "one news form today that's easy to miss is that we (OpenAI) again paused all big RL runs last Sunday because our newest model found a new loophole in our RL sandboxing that gave it live Internet access"
Read the full thread on Hacker News →
Related stories
- The Verge · 0 points · 4 days ago
- Hacker News · 55 points · 4 days ago
- OpenAI says planned GPT-6.1 is too insecure to releasearstechnica.comArs Technica · 0 points · 1 day ago
- The Verge · 0 points · 1 day ago
- Hacker News · 6 points · 4 days ago
- The Verge · 0 points · 8 days ago