Update README.md

README.md

@@ -17,8 +17,7 @@ pipeline_tag: feature-extraction
 
 # GeneJEPA: A Predictive World Model of the Transcriptome
 
-**GeneJEPA** is a Joint-Embedding Predictive Architecture (JEPA) trained for self-supervised representation learning on scRNA-seq.  
-It uses a Perceiver-style encoder to handle sparse, high-dimensional gene count vectors and learns from masked block prediction.
+**GeneJEPA** is a Joint-Embedding Predictive Architecture (JEPA) trained for self-supervised representation learning on scRNA-seq. It uses a Perceiver-style encoder to handle sparse, high-dimensional gene count vectors and a Fourier-feature tokenizer for numerical tokenization.
 
 **Why?** Produce compact cell embeddings you can use for clustering, transfer learning, linear probes, and downstream biological tasks.