The Topological Trouble With Transformers

Michael C. Mozer, Shoaib Ahmed Siddiqui, Rosanne Liu · 2026

Read The Topological Trouble With Transformers by Michael C. Mozer, Shoaib Ahmed Siddiqui, Rosanne Liu (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-2604-17121.txt.

Description

Mozer, Siddiqui, and Liu argue feedforward transformers fundamentally limit dynamic state tracking, pushing evolving state deeper into the layer stack with each input step until depth runs out. Dynamic-depth models and explicit or latent thinking can bypass the limit, but at high compute and memory cost. The authors call for recurrent architectures and implicit activation flows instead, offering a map of recurrent and continuous-thought designs by recurrence axis and token-to-step ratio.

Catalog record

CREATORMichael C. Mozer, Shoaib Ahmed Siddiqui, Rosanne Liu
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE32566
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.17121

Browse the full library · About the Tower of Babel