Metraipolis › archive › dc_dc3882b9ec6a

The Depth Flow of Token Representations Is Nonlinear and Does Not Descend Its Own Density — arXiv:2608.29706

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.29706
License (as archived)
CC0-1.0
Archived
2026-10-17T08:07:07.547Z
SHA-256 of the archived text
bdd38a2985b880f960b5a2ba6e3ed2cffb788fb0d6d0ddde10abc5822028c198
Stable URL of this page
https://metraipolis.com/documento/dc_dc3882b9ec6a

How to cite

The Depth Flow of Token Representations Is Nonlinear and Does Not Descend Its Own Density — arXiv:2608.29706 — arXiv (en), https://arxiv.org/abs/2608.29706 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_dc3882b9ec6a (Metraipolis archive dc_dc3882b9ec6a, sha256 bdd38a2985b880f960b5a2ba6e3ed2cffb788fb0d6d0ddde10abc5822028c198)
BibTeX
@misc{dcdc3882b9ec6a,
  title = {The Depth Flow of Token Representations Is Nonlinear and Does Not Descend Its Own Density — arXiv:2608.29706},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.29706},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_dc3882b9ec6a; sha256 bdd38a2985b880f960b5a2ba6e3ed2cffb788fb0d6d0ddde10abc5822028c198},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.29706. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A token's representation is carried through the network layer by layer. The whole vocabulary carried together forms a flow. We fit this flow's equation of motion as a discrete Langevin model over corpus-mean trajectories of Pythia-160M and Pythia-410M, and score the predicted steps on held-out tokens. Linear maps are often used as cheap surrogates for a layer. The flow they summarize is not linear: a quadratic drift beats the linear linear map at every transition of both models, and the Kramers--Moyal estimator agrees wherever its neighborhoods stay local. We then characterize the flow further. First, we show that it does not descend its own log-density. The drift instead descends a potential that is not the density. Second, the rotational component is not negligible, $4$ to $45\%$ of the explainable drift, and the circulation shows in what the flow preserves: a token keeps its angular rank across all thirteen layers while its norm rank is shuffled and its concentration rank is reversed by the last block.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-01 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2608.29706 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-17T07:47:52.728Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).