Skip to content

Tools Picking Hammer

pendingeasytools_picking——unowned

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

TrialStartedAgent timeModel requestsTokens in / outEst. costGrade
U✗ 1h 30m09-28 02:161h 30m25434.3M / 131k$0.443grader_error 0, missing_trajectory 1
L✗ 19m09-29 06:1819m533.0M / 51k$0.067deterministic 1, grader_error 0, live_success 0, n_actions 260, replay_success 0, video_rendered 1

Task instruction (upstream)

Select the blue hammer and put it on the table

tools_picking scene
No oracle video published upstream — this is the scene it starts from.
What RoboLab states about this task
Success predicate object_outside_of_and_on_surface
Predicate arguments object: ['blue_hammer'] · container: left_bin · surface: table · logical: all · require_gripper_detached: True
Scored subtasks 1
Subtask predicates pick_and_place_on_surface
Objects clamp cordless_drill spring_clamp husky_hammer blue_hammer red_hammer wood_hammer left_bin center_bin right_bin clamp_01
Upstream attributes semantics spatial color
Upstream difficulty simple
Episode budget 60 s
Other wordings vague — Get the blue hammer
specific — Reach into the tool area, grasp the blue-handled hammer, and place it flat on the table surface
Environment class ToolsPickingHammerTask
Upstream name ToolsPickingHammerTask
Defined in robolab/tasks/benchmark/tools_picking_hammer.py

From https://github.com/NVlabs/RoboLab @ v0.3.1.

Tags

Task DomainManipulation

Skill primitives in the demo: colorsemanticsspatial

Why this task is interesting

Not yet written.

Capability notes

Not yet written.

Oracle demo review

Not yet reviewed.

Discussion