Fix bibtex formatting, remove benchmark results section and benchmarking code link

README.md

@@ -17,7 +17,7 @@ size_categories:
   - 100K<n<1M
 ---
 
-# HUGO-Bench 
+# HUGO-Bench
 
 **Hierarchical Unsupervised Grouping of Organisms Benchmark**
 
@@ -179,23 +179,11 @@ mammals_data = dataset['mammals']
 
 **Note**: "Uncertain" indicates cases where manual validation was uncertain of the initial label.
 
-## Benchmark Results
-
-Using HUGO-Bench, we evaluated combinations of 5 ViT models  5 dimensionality reduction methods  4 clustering algorithms. Key findings:
-
-| Configuration | V-Measure | Notes |
-|---------------|-----------|-------|
-| DINOv3 + t-SNE + Hierarchical (K=30) | **0.958** | Best supervised |
-| DINOv3 + t-SNE + HDBSCAN | **0.943** | Best unsupervised |
-| DINOv3 + t-SNE + HDBSCAN | 33-37 clusters | ~10% overestimate of true species count |
-
-For complete benchmarking results, see our paper and the [interactive demo](https://hugomarkoff.github.io/animal_visual_transformer/).
-
 ## Citation
 
 If you use HUGO-Bench, please cite:
 
-`ibtex
+```bibtex
 @dataset{hugo_bench_2026,
   author       = {Markoff, Hugo and Bengtson, Stefan Hein and Ørsted, Michael},
   title        = {HUGO-Bench: Hierarchical Unsupervised Grouping of Organisms Benchmark},
@@ -204,12 +192,11 @@ If you use HUGO-Bench, please cite:
   url          = {https://huggingface.co/datasets/AI-EcoNet/HUGO-Bench},
   note         = {Expert-validated camera-trap images for zero-shot clustering. Image sources from LILA BC.}
 }
-`
+```
 
 ## Links
 
 -  [Interactive Demo](https://hugomarkoff.github.io/animal_visual_transformer/)
--  [Benchmarking Code](https://github.com/hugomarkoff/animal_visual_transformer)
 -  Paper: *Vision Transformers for Zero-Shot Clustering of Animal Images: A Comparative Benchmarking Study*
 
 ## Acknowledgments