Unverified50% confidenceBenchmarkExact time
自2023年GPT-4发布以来，AI在HumanEval等主流编程基准测试上的通过率从不足50%跃升至90%以上
1
Sources
50%
Confidence
Long-term
Relevance
7/23/2026
First Seen
Sources
Related Claims
UnverifiedThe pass rate of top models in 2025 on the HumanEval benchmark has exceeded 90%.76% similarVerifiedPass rates of code generation models on benchmarks such as HumanEval jumped from 60% to over 90% between 2024 and 202574% similarUnverified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%73% similarUnverified2023年人类在GAIA测试中的平均成功率约为90%，而当时最强的GPT-4在Level 1上仅勉强达到15%72% similarUnverifiedARC-AGI-2由AI安全研究员François Chollet设计，早期GPT-4在该测试上得分不足10%，人类平均得分约85%71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/596316API
curl https://kongchang.com/api/v1/knowledge/claims/596316MCP
get_claim(id=596316)