metadata
datasets:
- behavior-1k/2025-challenge-demos
- IliaLarchenko/behavior_224_rgb
license: apache-2.0
pipeline_tag: robotics
library_name: transformers
tags:
- behavior
- challenge
Task adaptation of Vision-Language-Action model: 1st Place Solution for the 2025 BEHAVIOR Challenge
This repository contains model checkpoints of the 1st place solution for the 2025 BEHAVIOR challenge.
For full documentation and code, refer to the GitHub repository.
Our technical report provides more details on the solution.
This model achieved ~26% final score on both Public and Private eval datasets.
The repo contains 4 checkpoints that are supposed to be used for different tasks:
| Checkpoint | Task Count | Task IDs |
|---|---|---|
| Checkpoint 1 | 20 tasks | 2, 3, 5, 6, 10, 11, 13, 14, 15, 19, 23, 24, 25, 28, 29, 34, 42, 44, 47, 48 |
| Checkpoint 2 | 16 tasks | 0, 1, 7, 8, 9, 12, 16, 17, 18, 20, 21, 22, 26, 30, 43, 45 |
| Checkpoint 3 | 13 tasks | 4, 27, 31, 32, 33, 35, 36, 37, 38, 39, 41, 46, 49 |
| Checkpoint 4 | 1 task | 40 |
Citation
If you find this work useful, please cite:
@misc{larchenko2025behavior,
title={Task adaptation of Vision-Language-Action model: 1st Place Solution for the 2025 BEHAVIOR Challenge},
author={Ilia Larchenko and Gleb Zarin and Akash Karnatak},
year={2025},
eprint={2512.06951},
archivePrefix={arXiv},
primaryClass={cs.RO},
url={https://arxiv.org/abs/2512.06951},
}