Unverified60% confidenceFactExact time
Agents' Last Exam is an evaluation system covering 55 professional domains that simulates long-running professional workflows.
2
Sources
60%
Confidence
Long-term
Relevance
7/13/2026
First Seen
Sources
Related Claims
UnverifiedAgent's Last Exam 是专为评估大模型自主代理能力设计的高难度基准,要求多步骤任务中持续调用外部工具、维持长上下文记忆并自我修正79% similarUnverifiedAgent's Last Exam 是由 Scale AI 与学术界联合设计的 Agent 能力基准测试75% similarVerifiedThe five-dimensional metric framework for Agent evaluation includes: integrity (task success rate), speed (execution efficiency), economy (cost control), stability (reliability), and safety (risk prevention and control).69% similarUnverifiedThe AB-620 exam covers three major domains: planning and configuring Agent solutions (30-35%), integrating and extending Agents (40-45%), and testing and managing Agents (20-25%).69% similarUnverifiedAgentPaySec categorizes test dimensions into six directions: authorization control, prompt injection, payment integrity, transaction tampering, tool abuse, and business logic vulnerabilities.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/497875API
curl https://kongchang.com/api/v1/knowledge/claims/497875MCP
get_claim(id=497875)