SENTIENT
Open Dashboard
Community registry online

Discover RL envs & benchmarks for every usecase.

Browse registry datasets, reusable benchmark suites, and environment templates for evaluating and improving agents across domains. No login required to explore.

Registry datasets

Public benchmarks available to inspect and run.

Benchmark tasks

Tasks ready for eval runs, regression gates, and grader tuning.

RL environments
Open

Environment templates and benchmarks for tool use, SWE, browsing, and agent recovery.

registry.sentient.devDATASET REGISTRY
Loading registry datasets…
Read-only for guests. Sign in to run and publish benchmarks.Open Community