Repository
Models / Research
Anchor Align · dwipddalal
Anchor-Align studies VLA fine-tuning with representation anchoring and language-action alignment.
Overview
Anchor-Align studies VLA fine-tuning with representation anchoring and language-action alignment.
Publisher summary: “Anchor-Align (arXiv:2607.13429): VLA finetuning that prevents behavior cloning from erasing pretrained VLM representations (catastrophic forgetting) and aligns language with actions. OOD generalization on a…”
Use its repository to inspect the implementation, documentation and supported workflows. This directory record identifies the maintainer, code license and related resources; it does not claim that the project has been installed or tested on your robot.
Project information
- Maintainer
- dwipddalal
- Primary language
- Python
- Repository code license
- MIT
- Model-weight license
- Not recorded
- Dataset license
- Not recorded
- Hardware design license
- Not recorded
- Maintenance
- Not archived in the last recorded metadata
- Last observed repository push
- 2026-08-28
Code licenses do not automatically cover model weights, datasets or hardware designs. Unrecorded license details require confirmation with the publisher.
Documentation sections
- Generalizable VLA Finetuning via Representation Anchoring and Language-Action Alignment
- Table of Contents
- Motivation
- Failure mode 1 — the pretrained VLM's semantics are silently erased
- Failure mode 2 — the model's language and actions disagree on the same observation
- Method: Anchor-Align
Browse this resource’s categories
Explore related collections
Vision-language-action models ↗More in Models
Repository
Repository
Research
starVLA · starVLA
StarVLA provides modular code for developing vision-language-action models.
Robot learning frameworks
Website
Research
OpenVLA-OFT
Fine-tuning tools for adapting OpenVLA action prediction.
Vision-language-action