Dominos¶
Agent runs
Each mode of this task runs once per run. Est. cost is tokens at list price (data/prices.yml), never a bill; see the RoboWits runs for every task of a run.
Codex + GPT-6 Luna, reasoning xhigh (ChatGPT login)default run
| Trial | Started | Agent time | Model requests | Tokens in / out | Est. cost | Grade |
|---|---|---|---|---|---|---|
| U✓ 11m | 09-26 01:42 | 11m | 33 | 1.1M / 17k | $0.026 | deterministic 1, grader_error 0, instance_exact 1, instance_ok 1, n_actions 336, video_rendered 1 |
| L✓ 1h 08m | 09-29 20:45 | 1h 08m | 150 | 15.8M / 114k | $0.240 | deterministic 1, grader_error 0, live_success 1, n_actions 1030, replay_success 1, video_rendered 1 |
Task instruction (upstream)
Push the red block onto the green target area.
Playback speed
What RoboWits states about this task
| Success criteria | 1. Red block's 2D footprint overlaps target area by at least 50% |
| Objects | red block target area white block 1 white block 2 white block 3 white block 4 |
| Physics | rigid |
| Episode budget | 200 control steps |
| Default control mode | EE_ABS |
| Evaluation scenes | 50 |
| Published mutations | 6 |
| Registry id | robowits/06-dominos-v0 |
| Environment class | DominosEnv |
| Defined in | gs_gym/envs/robowits/06_dominos.py |
From https://github.com/UMass-Embodied-AGI/RoboWits @ 9cc30ae.
Tags¶
Task DomainManipulation
Why this task is interesting¶
Not yet written.
Capability notes¶
Not yet written.
Oracle demo review¶
Not yet reviewed.