Update README.md

README.md

@@ -411,7 +411,7 @@ dataset_info:
 
 # NOLLI
 
-**NOLLI** is a benchmark of rule-based logical puzzles in **Korean and English**, designed with
+**NOLLI (논리)** is a benchmark of rule-based logical puzzles in **Korean and English**, designed with
 difficulty calibration in mind. It covers **15 task types x 3 difficulty tiers (easy / medium / hard)
 x 100 examples** each, for a total of **7,500 examples** (EN 3,000 + KO 4,500).