Add link to paper

README.md

@@ -1,25 +1,27 @@
 ---
+language:
+- en
 license: apache-2.0
+size_categories:
+- 1K<n<10K
 task_categories:
-  - question-answering
-  - text-retrieval
-language:
-  - en
+- question-answering
+- text-retrieval
 tags:
-  - rag
-  - retrieval-augmented-generation
-  - multi-hop-reasoning
-  - hotpotqa
-  - information-retrieval
-  - question-answering
-  - evaluation
-size_categories:
-  - 1K<n<10K
+- rag
+- retrieval-augmented-generation
+- multi-hop-reasoning
+- hotpotqa
+- information-retrieval
+- question-answering
+- evaluation
 ---
 
 # StratRAG
 
-**StratRAG** is a retrieval evaluation dataset for benchmarking Retrieval-Augmented Generation (RAG) systems on multi-hop reasoning tasks. It is derived from [HotpotQA](https://hotpotqa.github.io/) (distractor setting) and structured specifically for evaluating retrieval strategies — including sparse (BM25), dense, and hybrid approaches — in realistic, noisy document pool conditions.
+**StratRAG** is a retrieval evaluation dataset for benchmarking Retrieval-Augmented Generation (RAG) systems on multi-hop reasoning tasks. It was introduced in the paper [StratRAG: A Multi-Hop Retrieval Evaluation Dataset for Retrieval-Augmented Generation Systems](https://huggingface.co/papers/2604.22757).
+
+It is derived from [HotpotQA](https://hotpotqa.github.io/) (distractor setting) and structured specifically for evaluating retrieval strategies — including sparse (BM25), dense, and hybrid approaches — in realistic, noisy document pool conditions.
 
 ---
 
@@ -95,7 +97,8 @@ print("Number of docs in pool:", len(row["doc_pool"]))
 
 # Access gold documents directly
 for idx in row["gold_doc_indices"]:
-    print(f"\nGold doc [{idx}]:", row["doc_pool"][idx]["text"][:200])
+    print(f"
+Gold doc [{idx}]:", row["doc_pool"][idx]["text"][:200])
 ```
 
 ---
@@ -157,9 +160,6 @@ from datasets import load_dataset
 hotpot = load_dataset("hotpot_qa", "distractor")
 ```
 
----
-
-
 ---
 
 ## Benchmark Results