Create dataset card, add paper and GitHub links

README.md

@@ -0,0 +1,61 @@
+---
+task_categories:
+- other
+language:
+- en
+tags:
+- agent
+- tool-use
+- multi-hop-qa
+---
+
+# DeepSearch-World
+
+This repository contains the dataset presented in the paper [DeepSearch-World: Self-Distillation for Deep Search Agents in a Verifiable Environment](https://huggingface.co/papers/2607.07820). 
+
+[**GitHub Repository**](https://github.com/ornamentt/DeepSearch-World) | [**Environment Artifacts**](https://huggingface.co/datasets/Ornamentt/DeepSearch-World-Env)
+
+## Dataset Description
+
+DeepSearch-World contains a 420K multi-hop QA training pool and a verification set (DeepSearch-Val, provided in the `test` split) built from Wikipedia. It is designed to train and evaluate deep search agents that utilize search and page-reading tools to solve complex multi-step reasoning tasks.
+
+## Data Format
+
+Each question sample in the dataset contains at least the following fields:
+
+```json
+{
+  "id": 0,
+  "prompt": "Were Scott Derrickson and Ed Wood of the same nationality?",
+  "answers": ["yes"],
+  "question": "Were Scott Derrickson and Ed Wood of the same nationality?",
+  "answer": "yes"
+}
+```
+
+For scaffold teacher entity tracking, optional fields can be provided:
+
+```json
+{
+  "entity_list": ["Entity A", "Entity B"],
+  "fuzzed_features": {
+    "Entity A exact clue": "fuzzy clue for Entity A"
+  }
+}
+```
+
+## Citation
+
+If you find this dataset useful for your research, please cite:
+
+```bibtex
+@misc{geng2026deepsearchworldselfdistillationdeepsearch,
+      title={DeepSearch-World: Self-Distillation for Deep Search Agents in a Verifiable Environment}, 
+      author={Xinyu Geng and Xuanhua He and Sixiang Chen and Yanjing Xiao and Fan Zhang and Shijue Huang and Haitao Mi and Zhenwen Liang and Tianqing Fang and Yi R. Fung},
+      year={2026},
+      eprint={2607.07820},
+      archivePrefix={arXiv},
+      primaryClass={cs.CL},
+      url={https://arxiv.org/abs/2607.07820}, 
+}
+```