Codesota · Reinforcement Learning · Continuous Control · MuJoCoTasks/Reinforcement Learning/Continuous Control
Continuous Control · benchmark dataset · 2012 · N/A

Multi-Joint dynamics with Contact.

Physics-based continuous control benchmark. Evaluated on 15 DMControl tasks; metric is mean normalized score (0=random, 1000=expert) at 1M environment steps.

Paper ↗Submit a result ↵
§ 01 · Leaderboard

Best published scores.

9 results indexed across 1 metric. Shaded row marks current SOTA; ties broken by submission date.

Primary
average-return · higher is better
average-return· primary
9 rows
#ModelOrgSubmittedPaper / codeaverage-return
01TD-MPC2 (317M params)OSSUC San DiegoMar 2026iclr-2024-paper960
02TD-MPC2 (19M params)OSSUC San DiegoMar 2026iclr-2024-paper953
03FOWMOSSCMUMar 2026neurips-2024-paper945
04BROOSSDeepMind / TU WarsawMar 2026icml-2024-paper941
05TD-MPC2 (5M params)OSSUC San DiegoMar 2026iclr-2024-paper929
06DreamerV3OSSGoogle DeepMindMar 2026arxiv-paper897
07TD-MPCOSSUC San DiegoMar 2026icml-2022-paper857
08DrQ-v2OSSNYU / GoogleMar 2026icml-2022-paper799
09SAC (state-based)OSSUC BerkeleyMar 2026icml-2018-paper777
Fig 2 · Rows sorted by score within each metric. Shaded row marks SOTA. Dates reflect model or paper release where available, otherwise the date Codesota accessed the source.
§ 03 · Progress

1 steps
of state of the art.

Each row below marks a model that broke the previous record on average-return. Intermediate submissions are kept in the leaderboard above; only SOTA-setting entries are re-listed here.

Higher scores win. Each subsequent entry improved upon the previous best.

SOTA line · average-return
  1. Mar 22, 2026TD-MPC2 (317M params)UC San Diego960
Fig 3 · SOTA-setting models only. 1 entries span Mar 2026 → Mar 2026.
§ 06 · Contribute

Have a score that beats
this table?

Submit a checkpoint and a reproduction script. We will run it, publish the score, and — if it takes the top — annotate the step on the progress chart with your name.

Submit a result ↵Read submission guide
What a submission needs
  • 01A public checkpoint or API endpoint
  • 02A reproduction script with frozen commit + seed
  • 03Declared evaluation environment (Python, deps)
  • 04One row per metric declared by this dataset
  • 05A contact so we can follow up on discrepancies