AgentNativeResearchLab/rl-research-envs
rl-research-envs: Harbor dataset on Hugging Face with 4 tasks. Four self-contained research tasks for coding agents, plus 48 full agent runs on them (3 agents × 4 tasks × 4 trials), with trajectories, submitted code, and verifier output. Each task gives the agent a starter solution.py, a small…
Tasks
- Rank a pool of anomaly-detection alerts under a fixed investigation budget of k=20 per period. Value is time-decaying, false alerts…
- Find the change points in a set of series when the number of change points is unknown, varies per series, and is sometimes zero.
- Online multivariate anomaly detection on an industrial-sensor-like stream that also contains benign concept drift. Flag the injected…
- Online change-point detection on a 24-channel grouped stream. A regime break begins as a correlation build-up inside a single group…
AgentNativeResearchLab/rl-research-envs on the Hugging Face Hub