Bananas In Bin One More¶
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✓ 18m | 09-27 18:01 | 18m | 53 | 2.2M / 15k | $0.039 | deterministic 1, grader_error 0, instance_exact 1, instance_ok 1, n_actions 153, video_rendered 1 |
| L✗ 1h 30m | 09-29 06:54 | 1h 30m | 153 | 16.7M / 110k | $0.248 | deterministic 1, grader_error 0, live_success 0, n_actions 1522, replay_success 0, video_rendered 1 |
Task instruction (upstream)
Put one (1) more bananas 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: ['banana', 'banana_01', 'banana_03'] · container: grey_bin · logical: choose · K: 1 · require_gripper_detached: True · require_contact_with: False |
| Scored subtasks | 1 |
| Subtask predicates | pick_and_place |
| Objects | banana banana_01 banana_02 banana_03 banana_04 grey_bin |
| Upstream attributes | semantics counting |
| Upstream difficulty | simple |
| Episode budget | 60 s |
| Other wordings | vague — Add a banana to the bin specific — Pick up one additional banana and place it in the grey bin alongside the ones already there |
| Environment class | BananasInBinOneMoreTask |
| Upstream name | BananasInBinOneMoreTask |
| Defined in | robolab/tasks/benchmark/bananas_in_bin_one_more.py |
From https://github.com/NVlabs/RoboLab @ v0.3.1.
Tags¶
Task DomainManipulation
Skill primitives in the demo: countingsemantics
Why this task is interesting¶
Not yet written.
Capability notes¶
Not yet written.
Oracle demo review¶
Not yet reviewed.