Unverified50% confidenceReleaseExact time
研究团队发布了配套的基准测试集SkillCascade-Bench
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/28/2026
First Seen
Valid until: 12/27/2026
Sources
Related Entities
Related Claims
UnverifiedSkillCascade-Bench包含213个经过验证的级联测试用例,覆盖多种智能体系统与应用领域77% similarUnverified研究团队开发了SkillCascade,一个自动化的多智能体红队测试框架75% similarUnverified智能体评估基准测试套件包括GAIA、AgentBench和SWE-bench70% similarUnverified实验测试横跨六个异构基准(heterogeneous benchmarks)70% similarUnverifiedEVERY团队发明了名为SABench(Senior Engineer Benchmark)的基准测试,要求模型对质量较差的代码库进行从头重写70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/955501API
curl https://kongchang.com/api/v1/knowledge/claims/955501MCP
get_claim(id=955501)