Unverified50% confidenceBenchmarkExact time
GPT-5.4 has an accuracy of 50.5%, while its paired always-predict-up baseline is 51.6%.
1
Sources
50%
Confidence
Long-term
Relevance
9/11/2026
First Seen
Sources
Related Entities
Related Claims
Unverified在GPT-4发布初期，主流模型在GPQA上的准确率约为35-40%，人类领域专家基准线约为65%，随机猜测为25%77% similarUnverifiedIn the Agents' Last Exam benchmark, GPT-5.6 Sol achieved a score of 53.6, which is 13.1 points higher than Claude Fable 5 using adaptive reasoning.75% similarUnverifiedThe locally run Gemma 26B achieved an accuracy of only 49.2% across 453 scored predictions, while its paired always-predict-up baseline had an accuracy of 53.4%.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/903348API
curl https://kongchang.com/api/v1/knowledge/claims/903348MCP
get_claim(id=903348)