Unverified60% confidenceFactTime unknown
顶级PhD专家在Troubleshooting Bench上的得分约为36%，GPT 5.5即时版的得分仅略低于此
2
Sources
60%
Confidence
Medium-term (~90 days)
Relevance
5/31/2026
First Seen
Valid until: 8/29/2026
Sources
GPT 5.5即时版深度解析：幻觉率腰斩背后的能力与安全博弈
bilibili英文白斑马
Related Entities
Related Claims
UnverifiedTerminal Bench 2.1测试中GPT-5.6 Ultra得分接近92%，Fable 5为88%，考虑误差范围基本持平75% similarUnverifiedFable 5在SWE Bench Pro编程测试上得分80.3分，而OpenAI的GPT 5.5仅为58.6分69% similarUnverifiedGLM 5.2在Post Train Bench测试中得分34.3分，领先GPT的28.4分，但落后于Opus的37.2分69% similarUnverified在Agent's Last Exam基准测试中，GPT-5.6 Soul（extra high档）得分近54%，Fable（max档）为45%69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/3401API
curl https://kongchang.com/api/v1/knowledge/claims/3401MCP
get_claim(id=3401)