AWhat changed
Released a multi-robot vision-language-action model and training pipeline combining fused pretrained visual features, a language backbone and tokenized robot actions, with efficient target-robot fine-tuning.
Scientific scope
Released a multi-robot vision-language-action model and training pipeline combining fused pretrained visual features, a language backbone and tokenized robot actions, with efficient target-robot fine-tuning. Prismatic, Llama2, DINOv2, SigLIP and OpenX data are inherited. This is an accessible integrated system/resource, not invention of VLA or LoRA. Selected task success does not certify safe unattended control or universal generalization.
Evidence
The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Prismatic, Llama2, DINOv2, SigLIP and OpenX data are inherited. This is an accessible integrated system/resource, not invention of VLA or LoRA. Selected task success does not certify safe unattended control or universal generalization.
Contributors
- Source-listed authors and technical contributors · SCIENCE: Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.
- OpenVLA: An Open-Source Vision-Language-Action Model
- Kim et al.: open vision-language-action training and adaptation resource — original project
Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.
Full scientific record · Browse contribution grades · Request a correction