Leaked slide shows OpenAI's upcoming cyber model hitting 77% on ExploitGym
daniel_mac8 · x · 2026-11-13
A leaked slide shows OpenAI's 'Upcoming Cyber Model' reaching 77.0% success on ExploitGym, versus 52.6% for GPT-6 Cyber and 42.4% for GPT-6 Astra — nearly 2x improvement by average output tokens. The reposter argues the gains will transfer to domains beyond cybersecurity. Unconfirmed officially.
More from Models
- Thinking Machines releases Inkling 2 Small weights on Hugging Face, up to 383B-612B — ariG23498 · 2026-11-13
- Researchers praise Kimi's new architecture for adopting attnres: "best taste" — SonglinYang4 · 2026-11-13
- Google Search rolls out Gemini 4 Hydrogen globally in its biggest model upgrade yet — koraykv · 2026-11-13
- Sandbox escape attempts plummet since Claude Mythos 5, but nobody knows why — peterwildeford · 2026-11-13
- New model released with 'strong' results, weights now available — clarejtbirch · 2026-11-13
- Thinking Machines releases Inkling 2 Small weights under Apache 2.0 on Hugging Face — miramurati · 2026-11-13