| 📄 | 01_cover.svg | 905.3 KB |
| 📄 | 02_why_it_matters.svg | 9.0 KB |
| 📄 | 03_sequence_evolution.svg | 774.4 KB |
| 📄 | 04_rnn_cnn_limits.svg | 662.4 KB |
| 📄 | 05_architecture_overview.svg | 297.0 KB |
| 📄 | 06_encoder_decoder.svg | 8.7 KB |
| 📄 | 07_scaled_dot_product.svg | 25.0 KB |
| 📄 | 08_multi_head_attention.svg | 198.0 KB |
| 📄 | 09_three_uses.svg | 10.5 KB |
| 📄 | 10_ffn_residual_layernorm.svg | 568.0 KB |
| 📄 | 11_positional_encoding.svg | 1.0 MB |
| 📄 | 12_complexity_table.svg | 7.6 KB |
| 📄 | 13_training_setup.svg | 8.8 KB |
| 📄 | 14_results.svg | 8.6 KB |
| 📄 | 15_ablations.svg | 7.9 KB |
| 📄 | 16_conclusion.svg | 752.4 KB |