Update README.md

README.md

@@ -94,11 +94,17 @@ The model was trained with cross entropy loss for 150,000 iterations with a batc
 There is a [decoder model](https://huggingface.co/entropy/roberta_zinc_decoder) trained to reconstruct 
 inputs from embeddings generated with this model
 
-#### Compression Head
+#### Compression Encoder
 
-There is a [compression head model](https://huggingface.co/entropy/roberta_zinc_compression_head) 
+There is a [compression encoder model](https://huggingface.co/entropy/roberta_zinc_compression_encoder) 
 trained to compress embeddings generated by this model from the native size of 768 to 
-smaller sizes (512, 256, 128, 64, 32).
+smaller sizes (512, 256, 128, 64, 32) while preserving cosine similarity between embeddings.
+
+#### Decomposer
+
+There is a [embedding decomposer model](https://huggingface.co/entropy/roberta_zinc_enamine_decomposer) 
+trained to "decompose" a roberta-zinc embedding into two building block embeddings from the Enamine 
+library.
 
 
 **BibTeX:**