新云社 update: OpenAI said its control room, operating alongside a misalignment monitoring system, can stop model output within 35 seconds of a decision to intervene and halt or pause a task. The company disclosed the 35-second window as part of its oversight setup for detecting and acting on misaligned model behavior.
OpenAI: Control Room Can Halt Model Output Within 35 Seconds of Intervention
Share on WeChat
Open WeChat → Scan → then tap "…" to send to a chat or Moments.
Tap "…" in the top-right corner to send to a chat or share to Moments.
Related News
1h ago
Tencent Hunyuan’s Yao Shunyu says next-gen model to scale up significantly
1h ago
SoftBank’s Son Considered $1T Bid for SK Hynix, No Longer Active
1h ago
Anthropic researcher claims Claude helped prove Riemann Hypothesis
1h ago
Trump calls for Jensen Huang’s return to Nvidia after China trip
2h ago
CertiK and Columbia Researchers Complete First Production-Grade Formal Verification of DeepSeek, FlashAttention AI Kernels
2h ago
Zhejiang University Team Achieves World’s First QRAM on Superconducting Chip




