CLARU
RLE-Benchgated

kinex (v0.6.0) vs kinex (v0.6.0-improve) vs codex · LIBERO Long · GPT-6 Astra / medium

An evaluation dataset comparing three agent variants (kinex v0.6.0, kinex v0.6.0-improve, and codex) on LIBERO Long tasks using GPT-6 Astra medium model, containing 45 planned episodes with success rate metrics.

Downloads3K
Episodes45

Why This Matters for Physical AI

This evaluation benchmark enables comparative analysis of different embodied AI agent architectures on standardized manipulation tasks, providing empirical evidence for improving robot learning and control policies.

Technical Profile

Task Types
manipulation
Episodes
45
Annotation Types
action_labelsreward_labels
Part of the LIBERO family

Community Signals

Top 25% by downloads

Access

View on HuggingFace

This dataset requires access approval from the author on HuggingFace.

Need custom physical AI data?

Claru builds purpose-built datasets for any environment applications with dense human annotations and quality assurance.

Request a Sample Pack

Related Datasets

LIBERO

A vision-language-action dataset for robot manipulation with discrete actions, organized into spatial, object, goal, and long-horizon task categories.

rgblanguage
9K downloadsOct 2026MIT

TCoT LIBERO RLDS

RLDS/TFDS training data containing four filtered successful-demonstration suites and terminal windows of failed/retry rollouts used by libero_10_end60_retry experiments.

864 downloadsAug 2026other

Superseded: narrow-box random waypoint probing

An incomplete 2026-10-13 collection that was retired before final publication. This dataset is deprecated and should not be used as it has been superseded by an improved version.

545 downloadsOct 2026

chocopan-t3-reverse-oracle-rlds-v3

Synthetic scripted-oracle demonstrations of reverse manipulation tasks in simulation, where objects are taken out of containers or off plates and placed back on tables, formatted in RLDS/TFDS layout compatible with OpenVLA-OFT.

rgbproprioceptionlanguage
424 downloadsSep 2026mit

LIBERO LeRobot v3.0

Four standard LIBERO demonstration suites packaged in LeRobot v3.0 format containing synchronized RGB observations, robot state, actions, and natural-language task metadata for robot-policy and world-action-model training.

rgbproprioceptionlanguage
372 downloadsSep 2026cc-by-4.0

LIBERO Spatial, re-rendered at 224x224 (Patch Policy layout)

500 demonstrations of LIBERO Spatial (10 tasks x 50 demos) re-rendered at 224x224 resolution and converted to the Patch Policy layout, with replayed MuJoCo states and fixed fixture placements.

rgbproprioception
244 downloadsSep 2026apache-2.0