README.md

---
license: mit
---

# **PALM: Progress-Aware Policy Learning via Affordance Reasoning for Long-Horizon Robotic Manipulation**
![Teaser](main.png)
______________________________________________________________________

## 💡 Introduction
<div align="center">

<a href="#"><img src="https://img.shields.io/badge/Project-Page-blue?style=flat-square&logo=googlechrome&logoColor=white" alt="Project Page"></a>
<a href="https://arxiv.org/abs/2601.07060"><img src="https://img.shields.io/badge/Arxiv-2601.07060-red?style=flat-square&logo=arxiv&logoColor=white" alt="Arxiv"></a>
<a href="https://github.com/PLAN-Lab/PALM"><img src="https://img.shields.io/badge/GitHub-Repo-181717?style=flat-square&logo=github&logoColor=white"></a>
</div>

**Yuanzhe Liu, Jingyuan Zhu, Yuchen Mo, Gen Li, Xu Cao, Jin Jin, Yifan Shen, Zhengyuan Li, Tianjiao Yu, Wenzhen Yuan, Fangqiang Ding, Ismini Lourentzou**

Recent advancements in vision-language-action (VLA) models have shown promise in robotic manipulation, yet they continue to struggle with long-horizon, multi-step tasks. Existing methods lack internal reasoning mechanisms that can identify task-relevant interaction cues or track progress within a subtask, leading to critical execution errors such as repeated actions, missed steps, and premature termination. To address these challenges, we introduce PALM, a VLA framework that structures policy learning around interaction-centric affordance reasoning and subtask progress cues. PALM distills complementary affordance representations that capture object relevance, contact geometry, spatial placements, and motion dynamics, and serve as task-relevant anchors for visuomotor control. To further stabilize long-horizon execution, PALM predicts continuous within-subtask progress, enabling seamless subtask transitions. Across extensive simulation and real-world experiments, PALM consistently outperforms baselines, achieving a 91.8 \% success rate on LIBERO-LONG, a 12.5 \% improvement in average length on CALVIN ABC → D , and a 2 x improvement over real-world baselines across three long-horizon generalization settings.