README.md

---
license: mit
library_name: transformers
pipeline_tag: image-text-to-text
base_model: Qwen/Qwen2.5-VL-3B-Instruct
---

# AtlasVA: Self-Evolving Visual Skill Memory for Teacher-Free VLM Agents

This repository contains the model weights for **AtlasVA**, as presented in the paper [AtlasVA: Self-Evolving Visual Skill Memory for Teacher-Free VLM Agents](https://huggingface.co/papers/2605.17933).

**AtlasVA** is a teacher-free visual skill memory framework designed for Vision-Language Model (VLM) agents. It organizes memory into three complementary layers: spatial heatmaps, visual exemplars, and symbolic text skills. By evolving danger and affinity atlases directly from trajectory statistics, AtlasVA provides dense, coordinate-aware guidance for reinforcement learning, unifying perception, memory, and optimization without external LLM supervision.

- **Project Page:** [https://wangpan-ustc.github.io/AtlasvaWeb/](https://wangpan-ustc.github.io/AtlasvaWeb/)
- **Repository:** [https://github.com/wangpan-ustc/AtlasVA](https://github.com/wangpan-ustc/AtlasVA)
- **Paper:** [https://huggingface.co/papers/2605.17933](https://huggingface.co/papers/2605.17933)

## Model Details
- **Base Model**: Qwen2.5-VL-3B-Instruct
- **Task**: Multimodal agentic decision making (Sokoban, FrozenLake, 3D navigation, robotic manipulation).
- **Memory Layers**: Spatial heatmaps, visual exemplars, and symbolic text skills.

## Citation

```bibtex
@article{wang2026atlasva,
  title={AtlasVA: Self-Evolving Visual Skill Memory for Teacher-Free VLM Agents},
  author={Wang, Pan and Hu, Yihao and Liu, Xiujin and Yang, Jingchu and Wang, Hang and Wen, Zhihao},
  journal={arXiv preprint arXiv:2605.17933},
  year={2026}
}
```