Probing the topology of the space of tokens with structured prompts

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Robinson, Michael, Dey, Sourya, Kushner, Taisa
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866915205371396096
author Robinson, Michael
Dey, Sourya
Kushner, Taisa
author_facet Robinson, Michael
Dey, Sourya
Kushner, Taisa
contents This article presents a general and flexible method for prompting a large language model (LLM) to reveal its (hidden) token input embedding up to homeomorphism. Moreover, this article provides strong theoretical justification -- a mathematical proof for generic LLMs -- for why this method should be expected to work. With this method in hand, we demonstrate its effectiveness by recovering the token subspace of Llemma-7B. The results of this paper apply not only to LLMs but also to general nonlinear autoregressive processes.
format Preprint
id arxiv_https___arxiv_org_abs_2503_15421
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Probing the topology of the space of tokens with structured prompts
Robinson, Michael
Dey, Sourya
Kushner, Taisa
Differential Geometry
Artificial Intelligence
53Z50, 58Z05
I.2.7
This article presents a general and flexible method for prompting a large language model (LLM) to reveal its (hidden) token input embedding up to homeomorphism. Moreover, this article provides strong theoretical justification -- a mathematical proof for generic LLMs -- for why this method should be expected to work. With this method in hand, we demonstrate its effectiveness by recovering the token subspace of Llemma-7B. The results of this paper apply not only to LLMs but also to general nonlinear autoregressive processes.
title Probing the topology of the space of tokens with structured prompts
topic Differential Geometry
Artificial Intelligence
53Z50, 58Z05
I.2.7
url https://arxiv.org/abs/2503.15421