⭐ Star
0%
MULTI-HEAD ATTENTION HEATMAP

Self-Attention Weight Matrix Lab

Type any sentence to compute token-to-token Query-Key dot products and visualize the resulting Softmax attention weight heatmaps across attention heads.

📝 Input Sequence

📊 Token-to-Token Attention Heatmap [Softmax(QK^T / √d_k)]