Add paper, project page, and code links to dataset card

README.md

@@ -1,13 +1,13 @@
 ---
+language:
+- en
 license: cc-by-4.0
+size_categories:
+- 1K<n<10K
 task_categories:
 - video-text-to-text
 - multiple-choice
-language:
-- en
 pretty_name: Meta-Video
-size_categories:
-- 1K<n<10K
 configs:
 - config_name: meta_v
   data_files:
@@ -17,6 +17,10 @@ configs:
 
 # Meta-Video (meta_v)
 
+**Paper:** [Meta-Video: Efficient Standardized Evaluation for Video Understanding](https://huggingface.co/papers/2610.18535)  
+**Project page:** [https://soumyasj.github.io/meta-video](https://soumyasj.github.io/meta-video)  
+**Code:** [https://github.com/soumyasj/lmms_eval_video](https://github.com/soumyasj/lmms_eval_video)
+
 Meta-Video is a 1,000-question evaluation set for video understanding, sampled from five existing benchmarks
 (Video-MME, LongVideoBench, LVBench, VideoMMMU, VSI-Bench) over 791 videos (all <= 60 min).
 It keeps only questions that need the video: every question that at least two of three models
@@ -77,4 +81,4 @@ then the macro average over the five sources (VideoMMMU = mean of its three trac
 
 The question lists are released under CC BY 4.0. The videos remain under the licenses and terms of
 their original benchmarks and sources (Video-MME, LongVideoBench, LVBench, VideoMMMU, VSI-Bench /
-ScanNet, ScanNet++, ARKitScenes); please follow those terms.
+ScanNet, ScanNet++, ARKitScenes); please follow those terms.