# McGill-NLP/agent-reward-bench

From the catalogue. Listed by one researcher from one source page.

- id: `mcgill-nlp-agent-reward-bench`
- kind: other
- robot or device: not stated (class: other)
- how collected: not stated
- size as stated: not stated
- hours: not stated | hours only claimed: none | episodes: not stated
- year: 2025
- organisation: McGill-NLP
- licence: not stated (class: not stated)
- access: open (class: open)
- link (dataset): https://huggingface.co/datasets/McGill-NLP/agent-reward-bench
- Hugging Face repository: `McGill-NLP/agent-reward-bench`, 23,726 downloads in 30 days, https://huggingface.co/datasets/McGill-NLP/agent-reward-bench
- note: AgentRewardBench 💾Code 📄Paper 🌐Website 🤗Dataset 💻Demo 🏆Leaderboard AgentRewardBench: Evaluating Automatic Evaluations of Web Agent TrajectoriesXing Han Lù, Amirhossein ...
- files: https://huggingface.co/datasets/McGill-NLP/agent-reward-bench/tree/main. It has no LeRobot metadata at its top level, so episodes are not marked there. If it keeps datasets in folders, this lists them: https://datasets.gurasees.com/episodes/McGill-NLP/agent-reward-bench.md

Confirm the licence at the link before relying on it.