察元AI能力广场 · 资讯

Terminal-Bench-Science: Evaluating AI agents on scientific research workflows

作者 matt_d

评估 AI 智能体在科研工作流程中表现的基准测试集 Terminal-Bench-Science

智能体编排 命令行 研究调研 科研科学 测试 联网搜索

官方入口

同类资讯推荐

• GPT-5.6 Sol can beat Factorio on peaceful mode —
• Gemini 4 Carbon is the new #1 on the Artificial Analysis Intelligence Index v5 —
• Claude Code VS Code extension with backdoor released on Cursor marketplace —
• Claude Code regularly sends each background agent’s transcript to its model —
• Backdoors in official Claude Code extensions —
• Backdoor in Claude Code for VS Code extension on the Cursor Marketplace —
• Launch HN: Simantic (YC F26) – Firmware Simulation for AI Agents —
• We ran AI agent swarms in Minecraft and Mindustry —

浏览全部资讯 · 返回察元AI首页