Update README.md

README.md

@@ -11,6 +11,8 @@ configs:
   data_files:
   - split: test
     path: test.parquet
+  - split: raw
+    path: raw.parquet
 ---
 # Dataset Card for 🏹 CUPID Benchmark (COLM 2025)
 
@@ -47,6 +49,8 @@ Every instance is built from:
  - **1 current session**: The user’s new request, plus the gold preference string.
  - **8 prior sessions**: Multi-turn dialogues (6.4 turns on avg.) that implicitly expose the user's contextual preference.
 
+The `raw` split includes the user personas, context factors, and interaction sessions used to create these evaluation instances.
+
 **License**: CC‑BY‑4.0 (free for commercial & research use with attribution).
 
 ### Dataset Structure