Geometric Analysis of Token Selection in Multi-Head Attention
Timur Mudarisov, Mikhail Burtsev, Tatiana Petrova, Radu State · 2026
Read Geometric Analysis of Token Selection in Multi-Head Attention by Timur Mudarisov, Mikhail Burtsev, Tatiana Petrova, Radu State (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-01893.txt.
Description
Timur Mudarisov and colleagues view multi head attention in language models through a top-N selection lens, studying conduct in value state space. They define Precision, Recall and F-score metrics for token selection, and derive bounds from dimension and margin under assumptions. Stable value norms with a sink token, exponential similarity decay, and attention weight profiles predict a small-N regime of strongest separability. Across three models results track envelopes, selection sharpens separability, sink similarity tracks Recall, and LLaMA-2-7B heads split into Retriever, Mixer and Reset regimes.
Catalog record
| CREATOR | Timur Mudarisov, Mikhail Burtsev, Tatiana Petrova, Radu State |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 40303 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602.01893