Update README.md

README.md

@@ -105,14 +105,16 @@ configs:
 
 The Multi-factor Sequential Disentanglement benchmark includes a **modified variant of the dSprites dataset**, adapted to support sequential multi-factor disentanglement.
 
-- Original repository:  
+- Original repository:
   https://github.com/deepmind/dsprites-dataset
-
-- Reference paper:  
-  M. Higgins, L. Matthey, A. Pal, C. Burgess, X. Glorot, M. Botvinick, S. Mohamed, A. Lerchner.  
-  *beta-VAE: Learning Basic Visual Concepts with a Constrained Variational Framework*, ICLR 2017.  
-  https://openreview.net/forum?id=Sy2fzU9gl
-
+```
+@misc{dsprites17,
+author = {Loic Matthey and Irina Higgins and Demis Hassabis and Alexander Lerchner},
+title = {dSprites: Disentanglement testing Sprites dataset},
+howpublished= {https://github.com/deepmind/dsprites-dataset/},
+year = "2017",
+}
+```
 ⚠ **Note:** The dSprites dataset is licensed under the Apache License 2.0. We redistribute it here solely for non-commercial research purposes, following the original license terms. Please cite the above paper when using this dataset in your work.
 
-In this modified sequential version, the object is colorized and remains static in position, while the dynamic factors are its **scale** and **orientation**, allowing targeted evaluation of disentanglement in temporal settings.
+In this modified sequential version, the object's **color**, **shape** and **position** are fixed, while the dynamic factors are its **scale** and **orientation**, allowing targeted evaluation of disentanglement in temporal settings.