Archive - Paulius Rauba

Why we divide by the square root of the key dimension in attention

There are three implicit assumptions we make

Oct 12, 2025 • Paulius Rauba

Attention mechanism for non-Euclidean spaces

Looking at the geometric assumptions behind attention.

Oct 5, 2025 • Paulius Rauba

#nojs-banner { position: fixed; bottom: 0; left: 0; padding: 16px 16px 16px 32px; width: 100%; box-sizing: border-box; background: red; color: white; font-family: -apple-system, "Segoe UI", Roboto, Helvetica, Arial, sans-serif, "Apple Color Emoji", "Segoe UI Emoji", "Segoe UI Symbol"; font-size: 13px; line-height: 13px; } #nojs-banner a { color: inherit; text-decoration: underline; } This site requires JavaScript to run correctly. Please turn on JavaScript or unblock scripts