Unverified50% confidenceFactExact time
During the suspension, the laboratory focused on three tasks: hardening the research environment, conducting red-team exercises, and expanding monitoring coverage.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/31/2026
First Seen
Valid until: 11/29/2026
Sources
Related Entities
Related Claims
Unverified该论文选择国际象棋作为实验平台，系统性研究预训练算力分配如何影响后训练阶段的下游任务表现，并完整复现了预训练、SFT、RL三个阶段57% similarUnverifiedRed-teaming originated in the military and cybersecurity fields and, in the AI safety context, has developed into multiple layers including prompt-level attacks, training infrastructure security testing, and model capability risk testing.56% similarUnverifiedA high-quality RL training environment must satisfy three conditions: providing a task distribution close to real-world scenarios, delivering accurate and discriminative reward signals, and supporting large-scale parallel sampling to accelerate training.56% similarUnverifiedRL3内置支持训练检查点保存与恢复、环境与模型分享、复用他人环境或策略等协作功能56% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/827367API
curl https://kongchang.com/api/v1/knowledge/claims/827367MCP
get_claim(id=827367)