Unverified50% confidenceFactExact time
Terminal-Bench-Science is a benchmark specifically designed to evaluate the performance of AI agents in real scientific research workflows.
1
Sources
50%
Confidence
Long-term
Relevance
9/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedTerminal-Bench-Science is better understood as a tool for measuring AI-assisted scientific research capabilities rather than a declaration that AI can already conduct research independently.79% similarUnverifiedTerminal Bench is an evaluation benchmark for measuring the real operational capabilities of large language models and AI Agents in command-line terminal environments79% similarVerified代理化编程强调测试驱动闭环，测试为AI提供明确的成功信号，是AI自我纠错的关键72% similarUnverifiedAI Skills can automatically infer parameter boundary values, generate positive and negative test cases, and output executable scripts conforming to Pytest or unittest specifications by parsing OpenAPI/Swagger documents or interface descriptions.71% similarUnverifiedMETR是专注于AI能力评估的研究机构，其发布的任务时长基准测试按人类完成所需时间对软件工程任务分级，让AI代理独立完成并统计成功率71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/912324API
curl https://kongchang.com/api/v1/knowledge/claims/912324MCP
get_claim(id=912324)