Pith. sign in

REVIEW 1 cited by

Towards Universality: Studying Mechanistic Similarity Across Language Model Architectures

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2410.06672 v2 pith:ZLM2KLWH submitted 2024-10-09 cs.CL

classification cs.CL
keywords modelssimilarsimilaritytransformersuniversalityarchitecturesfeatureslanguage
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

The hypothesis of Universality in interpretability suggests that different neural networks may converge to implement similar algorithms on similar tasks. In this work, we investigate two mainstream architectures for language modeling, namely Transformers and Mambas, to explore the extent of their mechanistic similarity. We propose to use Sparse Autoencoders (SAEs) to isolate interpretable features from these models and show that most features are similar in these two models. We also validate the correlation between feature similarity and Universality. We then delve into the circuit-level analysis of Mamba models and find that the induction circuits in Mamba are structurally analogous to those in Transformers. We also identify a nuanced difference we call \emph{Off-by-One motif}: The information of one token is written into the SSM state in its next position. Whilst interaction between tokens in Transformers does not exhibit such trend.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. What, Where, and How: Disentangling the Roles of Task, Language, and Model in Code Model Representations

    cs.CL 2026-07 conditional novelty 7.0 of 10

    Across Python and Rust, two code models conserve the ranking of which constructs get dedicated circuits (ρ≈0.65) but differ by 12-13 layers in where those circuits sit and whether they show an early spike.

Pith tools