精选理由
Agent Arena上线新基准，用因果追踪测模型在长期任务中的真实表现，还能按类别筛选对比。
Agent Arena使用因果追踪技术评估模型在现实世界长期任务上的表现。该基准测试显示模型相对于平均模型的净改进程度。用户可访问arena.ai/leaderboard查看整体排行榜，并按代码、聊天和工作类别进行筛选。
原文 · lmarena.ai
Agent Arena measures performance on real-world, long-horizon tasks using causal tracing. Net improve...
Agent Arena measures performance on real-world, long-horizon tasks using causal tracing. Net improvement shows how much a model improves outcomes relative to the average model. Dive into the Agent Arena leaderboard overall and now filter by categories Code, Chat and Work: arena.ai/leaderboard/ag… 💬 0 🔄 0 ❤️ 0 👀 783 ⚡