Unverified50% confidenceFactExact time
The PsyAgentBench study made public as many as 41,904 trial data records.
1
Sources
50%
Confidence
Long-term
Relevance
9/22/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe research team spent $42.57 on large-scale experiments across 23,440 scoring episodes.62% similarUnverified2023年加州大学伯克利分校发布的MT-Bench基准测试发现，顶级大模型作为评判者与人类专家判断一致性可达80%以上59% similarUnverified2024年一项研究对GPT-4在Temperature=0下的代码生成任务进行100次测试，发现约8%的测试用例存在跨次运行的输出差异58% similarUnverifiedDiscoBench数据集从影视、体育、游戏、学术、金融等11个真实领域收集多跳问答样本，最终生成211个测试样本，包含463个歧义实例56% similarPartially VerifiedIn the test case, the cache hit rate reached as high as 99.7%, with nearly 39 million tokens out of approximately 40 million tokens generated hitting the cache.55% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/939984API
curl https://kongchang.com/api/v1/knowledge/claims/939984MCP
get_claim(id=939984)