Add dataset card, paper link, and GitHub repository

README.md

@@ -0,0 +1,40 @@
+---
+task_categories:
+- robotics
+---
+
+# Non-Markovian Long-Horizon Robot Manipulation via Keyframe Chaining
+
+This repository contains the simulation-based datasets used in the paper [Non-Markovian Long-Horizon Robot Manipulation via Keyframe Chaining](https://huggingface.co/papers/2603.01465).
+
+[**Code**](https://github.com/cytoplastm/KC-VLA) | [**Paper**](https://huggingface.co/papers/2603.01465)
+
+## Description
+
+The dataset consists of custom-generated trajectories for four Non-Markovian manipulation tasks built upon the [ManiSkill](https://github.com/haosulab/ManiSkill) simulator. These tasks are specifically designed to measure task success rates in scenarios where optimal robot actions rely on specific past states (Non-Markovian dependencies) rather than just the current observation. 
+
+This data was used to train and evaluate the **Keyframe-Chaining VLA** framework, which utilizes a task-modulated Keyframe Selection Module (KSM) to efficiently ground long-horizon dependencies.
+
+## Dataset Structure
+
+The dataset includes trajectories for 4 custom-generated tasks:
+- Built on the ManiSkill simulation backbone.
+- Focused on long-horizon, memory-dependent manipulation.
+- Generated via a custom collection pipeline.
+
+## Usage
+
+Please refer to the [official GitHub repository](https://github.com/cytoplastm/KC-VLA) for instructions on environment setup and how to use this data within the KC-VLA training and evaluation pipeline.
+
+## Citation
+
+If you find this dataset or the custom ManiSkill benchmark useful for your research, please consider citing:
+
+```bibtex
+@article{KC-VLA,
+  title={Non-Markovian Long-Horizon Robot Manipulation via Keyframe Chaining},
+  author={Yipeng Chen and Wentao Tan and Lei Zhu and Fengling Li and Jingjing Li and Guoli Yang and Heng Tao Shen},
+  journal={arXiv preprint arXiv:2603.01465},
+  year={2026},
+}
+```