Toolathlon-Verified_Trajectories
收藏官方服务：
资源简介：
This dataset contains the full trajectories we collected in running Toolathlon-Verified Evaluation. Now contains: - Claude Opus 4.8 (max) - GPT 5.5 (xhigh) - Gemini 3.5 Flash (high) - GLM 5.2 (max) - Kimi K2.7 Code - DeepSeek V4 Pro (max)
提供机构：
maas创建时间：
2026-07-02
