Red Dishes In Bin¶
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 15m | 09-26 07:05 | 1h 15m | 208 | 24.1M / 101k | $0.322 | deterministic 1, grader_error 0, instance_exact 1, instance_ok 1, n_actions 527, video_rendered 1 |
| L✗ 1h 30m | 09-29 11:21 | 1h 30m | 243 | 31.5M / 175k | $0.468 | deterministic 1, grader_error 0, live_success 0, n_actions 4590, replay_success 0, video_rendered 1 |
Task instruction (upstream)
Put the red dishware in the grey bin
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: ['mug', 'bowl'] · container: grey_bin · logical: all · require_gripper_detached: True |
| Scored subtasks | 1 |
| Subtask predicates | pick_and_place |
| Objects | mug bowl grey_bin banana rubiks_cube |
| Upstream attributes | color semantics |
| Upstream difficulty | simple |
| Episode budget | 60 s |
| Other wordings | vague — Clean up red dishware specific — Pick up each red mug and bowl from the table and place them in the grey bin |
| Environment class | RedDishesInBinTask |
| Upstream name | RedDishesInBinTask |
| Defined in | robolab/tasks/benchmark/red_dishes_in_bin.py |
From https://github.com/NVlabs/RoboLab @ v0.3.1.
Tags¶
Task DomainManipulation
Skill primitives in the demo: colorsemantics
Why this task is interesting¶
Not yet written.
Capability notes¶
Not yet written.
Oracle demo review¶
Not yet reviewed.