added better table

README.md

@@ -27,12 +27,16 @@ We tested the speed of the indexing on affordable GPUs , we pass the embeddings
 | NVIDIA t4    | 4           | 4.55                |
 
 ### Interpretability 
-| Query          | Images |  
-|--------------|-------------|
-| Scaled  |<p align="center"><img src="images/hm_token_1.png" width="800px"></p>|
-| Dot    |<p align="center"><img src="images/hm_token_3.png" width="800px"></p> |
-| what is the model architecture and what is adaptive visual encoding?  |<p align="center"><img src="images/hm_token_map_32.png" width="800px"></p> |
-| what is the model architecture and what is adaptive visual encoding?  |<p align="center"><img src="images/hm_token_41.png" width="800px"></p> |
+| Query                                              |  Image                                               |  Image                                                  |
+|----------------------------------------------------|------------------------------------------------------------|------------------------------------------------------------|
+| Scaled and Dot                                     |<p align="center"><img src="images/hm_token_1.png" width="800px"></p>|<p align="center"><img src="images/hm_token_3.png" width="800px"></p>|
+| What is the model architecture and what is adaptive visual encoding?  |<p align="center"><img src="images/hm_token_map_32.png" width="800px"></p>|<p align="center"><img src="images/hm_token_41.png" width="800px"></p>|
+
+What does this heatmap tell us ?
+- The heatmap shows areas of high attention (bright spots) and low attention (darker areas) for a specific token.
+-  The model seems to understand and focus on the relevant parts of the image that discuss or illustrate the adaptive visual encoding concept.
+- The spread of attention indicates how precisely the model can identify the relevant areas. In this case, the attention seems to be spread across relevant diagrams and text, suggesting a good understanding.
+
 
 ## 📚 Resources