b:botstest.Submit a resource ↗

Models / Research

vla from scratch · aghasalim

VLA From Scratch compares action representations on multimodal demonstrations and obstacle-routing tasks.

action-chunkingdiffusion-policyflow-matchingfrom-scratchimitation-learning
Publisher repository ↗
vla from scratch · aghasalim resource page screenshot
Resource screenshot · captured 2026-10-08 · visually reviewed

Overview

VLA From Scratch compares action representations on multimodal demonstrations and obstacle-routing tasks.

Publisher summary: “Four VLA action representations compared on multimodal demonstrations: bins, regression, diffusion, flow. Regression drives into the obstacle it should route around; flow keeps multimodality…”

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
aghasalim
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
2026-10-02

Code licenses do not automatically cover model weights, datasets or hardware designs. Unrecorded license details require confirmation with the publisher.

Documentation sections

  • vla-from-scratch
  • Scope, stated first
  • The task, and why it has an obstacle
  • Results
  • Sampling error
  • Latency is where flow earns its place

Browse this resource’s categories

Models / Vision-language-action ↗

Explore related collections

Vision-language-action models ↗

More in Models