Expired50% confidenceFactTime unknown
Human experts score an average of approximately 55%–65% on the MMMU Pro benchmark.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
Valid until: 9/30/2026(expired)
Sources
Gemini 3 Flash In-Depth Review: Comprehensive Testing of Coding, Multimodal, and Writing Capabilities
bilibilikate人不错
Related Claims
Unverified斯坦福HAI研究所2023年报告显示，最新LLM在HumanEval等代码基准测试中正确率已超过85%，远超2021年的不足50%68% similarUnverifiedThe system achieved a detection accuracy rate of over 85% on a validated sample of 50 CVEs.66% similarUnverified人类领域专家在 GPQA 上平均约得 65%65% similarUnverifiedThe average human score on the ARC-AGI test is approximately 85%, while top AI models have long hovered in the 30%–60% range.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/49607API
curl https://kongchang.com/api/v1/knowledge/claims/49607MCP
get_claim(id=49607)