AgentRoboCritic-Muse-Glimmer-30B

Papers and technical reports

AgentRoboCritic is a vision-language critic for executed robot manipulation actions. It uses the task, local intent, action command and images around execution to judge whether each action achieved its intent, categorize failures and explain the visual evidence. The released model fine-tunes Muse-Glimmer-30B.

View full model description

Evaluate individual executed robot manipulation actions against the stated local intent and diagnose failure mechanisms from visual observations.

Papers and technical reports

Paper / report unconfirmed

Official introductory materials are recorded; no version-specific paper or technical report has been verified.

Last checked: 2026-11-20

How to access and use

Model files
Public access
Training data
Public access

Training code for this version is unconfirmed. Public code resources do not establish access to the exact training implementation. View verification notes

Training / fine-tuning code

Training data

Some training datasets are public; this does not establish access to the complete training mixture.

Model files

Model files
AgentRoboCritic-Muse-Glimmer-30B (BF16)AccessWeights licenseCommercial use
ConditionalTerms

SourceVerified on · 2026-11-20

Conditions

Apache 2.0 as declared by the publisher, subject to the Muse Glimmer Usage Policy.

Release information

SourceVerified on · 2026-11-20

Checks before deployment

The reported evaluation covers 400 human-labeled held-out manipulation actions. The system message allows inconclusive judgments when observations are insufficient. Available weights or code do not establish plug-and-play robot compatibility. Check the required configuration, data and interfaces in the source documentation.

Deployment and usage resources

Runtime and frameworks
Transformers

Architecture and inputs / outputs

Model type
Reward & evaluation
Weight parameters
29.7766 B
Input
Language / RGB images
Released
November 20, 2026

Scope and training

Base model

meta-models/Muse-Glimmer-30B

Pipeline task

image-text-to-text

Prediction output

JSON judgment of intent achievement, failure mechanism, grasp failure subcategory and visually grounded reason

Training data scale

12844 actions

Training data sources

https://huggingface.co/datasets/cx-cmu/AgentRoboCritic-Dataset

Lineage

Fine-tuned from: meta-models/Muse-Glimmer-30B ↗

Training data

  • AgentRoboCritic-Bench

    EvaluationAvailable

    400 held-out robot manipulation actions annotated by humans, including 264 failed actions. CC BY 4.0 covers the annotations. Images remain subject to their source benchmarks’ licenses.

    400 actions · CC BY 4.0

  • AgentRoboCritic-Dataset

    Fine-tuningAvailable

    Supervised rank-32 LoRA was merged into the base weights. The authors report 12,844 training actions across 12 manipulation benchmarks, 2 epochs, batch size 32 and a cosine learning-rate schedule starting at 5×10⁻⁵. CC BY 4.0 covers the annotations. Images remain subject to their source benchmarks’ licenses.

    12844 actions · CC BY 4.0

Full parameters and verification notes (model card)

The Hugging Face publisher is Chenyan Xiong Research Group at CMU. Project author affiliations are Carnegie Mellon University, University of Illinois at Urbana-Champaign, and Bosch Research North America & Bosch Center for Artificial Intelligence (BCAI).

Supervised rank-32 LoRA was merged into the base weights. The authors report 12,844 training actions across 12 manipulation benchmarks, 2 epochs, batch size 32 and a cosine learning-rate schedule starting at 5×10⁻⁵.

The training and validation splits contain 12,844 and 1,548 actions, respectively. On the held-out evaluation the authors also report outcome Cohen’s κ of 0.444, failure-mode Cohen’s κ of 0.484, and an evaluation cost of USD 0.79.

Evaluation used temperature 0 and a maximum of 1,024 output tokens.

Submit a source or correction