Jugs On Shelf¶
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 RoboLab 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✗ 1h 00m | 09-26 02:02 | 1h 00m | 135 | 17.4M / 103k | $0.252 | grader_error 0, missing_trajectory 1 |
| L✗ 1h 30m | 09-29 10:18 | 1h 30m | 207 | 27.0M / 151k | $0.421 | deterministic 1, grader_error 0, live_success 0, n_actions 2889, replay_success 0, video_rendered 1 |
Task instruction (upstream)
Put all the jugs on the shelf
No oracle video published upstream — this is the scene it starts from.
What RoboLab states about this task
| Success predicate | object_in_container |
| Predicate arguments | object: ['utilityjug_a01', 'utilityjug_a02'] · container: large_storage_rack · logical: all · require_gripper_detached: True |
| Scored subtasks | 1 |
| Subtask predicates | pick_and_place |
| Objects | large_storage_rack whitepackerbottle_a01 whitepackerbottle_a02 utilityjug_a01 utilityjug_a02 squarepail_a01 plasticpail_a02 whitepackerbottle_a03 |
| Upstream attributes | semantics |
| Upstream difficulty | simple |
| Episode budget | 120 s |
| Other wordings | vague — Put the jugs on the shelf specific — Pick up both jugs from the table and place them on the shelf |
| Environment class | JugsOnShelfTask |
| Upstream name | JugsOnShelfTask |
| Defined in | robolab/tasks/benchmark/jugs_on_shelf_task.py |
From https://github.com/NVlabs/RoboLab @ v0.3.1.
Tags¶
Task DomainManipulation
Skill primitives in the demo: semantics
Why this task is interesting¶
Not yet written.
Capability notes¶
Not yet written.
Oracle demo review¶
Not yet reviewed.