README.md

metadata

title: ToolUse.1
emoji: 🧰
colorFrom: gray
colorTo: red
sdk: static
pinned: true
license: apache-2.0
short_description: Tool-calling leaderboard that scores abstention

ToolUse.1 — Leaderboard

Static leaderboard for the ToolUse.1 tool-calling benchmark. It renders results.json, which is generated from harness result files by publish.py in the benchmark repository.

Add a row by running the harness yourself and opening a discussion on the dataset repository with the raw results file.