Unverified50% confidenceBenchmarkExact time
Human test subjects achieve an average accuracy of over 85% on ARC problems, while mainstream large language models have long hovered below 20%.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/11/2026
First Seen
Valid until: 1/9/2027
Sources
Related Entities
Related Claims
Unverified顶尖大语言模型在HumanEval基准测试上的Pass@1准确率已超过85%73% similarUnverifiedThe composite score in this study places 80% weight on the large language model's judgment of factual entailment and 20% on lexical overlap.71% similarUnverifiedResearch shows that the agreement rate between LLM judges and human scores is approximately 80%, approaching but not yet surpassing the level of human inter-annotator agreement (approximately 81%).68% similarUnverifiedAI detection methods based on text statistical features achieve an accuracy of approximately 70%–85% under ideal conditions, but effectiveness drops significantly after text has been manually modified.68% similarVerifiedIn a well-calibrated model, an output score of 0.8 should statistically correspond to an actual accuracy of 80%.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/996000API
curl https://kongchang.com/api/v1/knowledge/claims/996000MCP
get_claim(id=996000)