datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
dojo-bench-mini
Dojo-Bench-Mini
Dojo-Bench-Mini is a small public bench containing tasks for running computer use agents against "mocked" productivity software and games. These include:
Linear
LinkedIn
Tic-Tac-Toe
2048
For full details on running this benchmark, check-out:
Dojo, an environment hub for computer use agents
Docs
Running an evaluation
jailbreak-dojo-corpusjailbreak-dojo-leaderboard