Unverified50% confidenceFactExact time
ECI aims to address the problem of traditional benchmarks (such as MMLU, HumanEval, etc.) operating in silos and being difficult to compare horizontally, answering the question of what point in time a model's capabilities are equivalent to.
1
Sources
50%
Confidence
Long-term
Relevance
9/4/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAISI的评估采用多层次测试框架，参照CVE数据库中的真实案例，同时设计全新挑战场景测试模型泛化能力71% similarUnverifiedAISI的评估方法包括标准化的CTF风格挑战题目、真实世界漏洞的复现任务、以及模拟渗透测试场景69% similarUnverifiedAISI的网络安全评估采用CTF挑战题、真实世界CVE漏洞复现测试、以及受控环境中对模拟系统的自主渗透测试等多层次测试方法68% similarUnverifiedECC adopts a research-first development methodology, where agents are given ample time to research context, consult materials, and understand the nature of a problem before entering the implementation phase.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/855557API
curl https://kongchang.com/api/v1/knowledge/claims/855557MCP
get_claim(id=855557)