Papers › Not All Language Model Features Are Linear

Not All Language Model Features Are Linear

23 May 2024arXiv:2405.14860archive 2025-07-28

Joshua Engels, Eric J. Michaud, Isaac Liao, Wes Gurnee, Max Tegmark

Recent work has proposed that language models perform computation by manipulating one-dimensional representations of concepts ("features") in activation space. In contrast, we explore whether some language model representations may be inherently multi-dimensional. We begin by developing a rigorous definition of irreducible multi-dimensional features based on whether they can be decomposed into either independent or non-co-occurring lower-dimensional features. Motivated by these definitions, we design a scalable method that uses sparse autoencoders to automatically find multi-dimensional features in GPT-2 and Mistral 7B. These auto-discovered features include strikingly interpretable examples, e.g. circular features representing days of the week and months of the year. We identify tasks where these exact circles are used to solve computational problems involving modular arithmetic in days of the week and months of the year. Next, we provide evidence that these circular features are indeed the fundamental unit of computation in these tasks with intervention experiments on Mistral 7B and Llama 3 8B. Finally, we find further circular representations by breaking down the hidden states for these tasks into interpretable components, and we examine the continuity of the days of the week feature in Mistral 7B.

PaperPDFCodeCode Syntology ran

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2405.14860")

Code

Syntology Ran 6 of 10 code samples harvested from 1 repository linked to this paper; 4 have no recorded run. Of those that ran: 6 ran with no contract checked.

By repository: official repository: 10 samples from 1 repository, 6 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

joshengels/multidimensionalfeatures officialmentioned in papermentioned on GitHubpytorchMIT report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

10 samples harvested; 6 ran; 0 honoured the contract we drafted; 4 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

6ran
4unverified

Licence: 0 of the 10 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from joshengels/multidimensionalfeatures. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples from a repository marked official are listed first. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

expand_descriptions joshengels/multidimensionalfeatures/intervention/utils.py official repository ran MIT (permissive) · e5bbb58953bc029e · report
f joshengels/multidimensionalfeatures/feature_deconstruction/simple_example.py official repository ran MIT (permissive) · 35bb96ea1f68e1fc · report
get_all_acts joshengels/multidimensionalfeatures/intervention/task.py official repository ran MIT (permissive) · 61e1dc25eff28dd1 · report
get_lambdas_and_descriptions joshengels/multidimensionalfeatures/intervention/utils.py official repository ran MIT (permissive) · 0400998cc6c8e217 · report
get_points joshengels/multidimensionalfeatures/intervention/intervene_in_middle_of_circle.py official repository ran MIT (permissive) · dc43373d010228a4 · report
mean_confidence_interval joshengels/multidimensionalfeatures/intervention/compare_circle_intervention_types.py official repository ran MIT (permissive) · 6661fded31c6734f · report
do_regression joshengels/multidimensionalfeatures/intervention/circle_finding_utils.py official repository unverified MIT (permissive) · 6d8c73b8de5b9b2c · report
find_c_circle joshengels/multidimensionalfeatures/intervention/circle_finding_utils.py official repository unverified MIT (permissive) · fb6d83ae4739bc75 · report
get_acts joshengels/multidimensionalfeatures/intervention/task.py official repository unverified MIT (permissive) · 5daefa372b9d0d79 · report
get_acts_pca joshengels/multidimensionalfeatures/intervention/task.py official repository unverified MIT (permissive) · 08d36115ab4d9a96 · report

Tasks

AllLanguage ModelingLanguage Modellingmodel

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

AdamAttentionAttention DropoutBPECosine AnnealingDense ConnectionsDiscriminative Fine-TuningDropoutGPT-2LLaMALayer NormalizationLinear LayerLinear Warmup With Cosine AnnealingMulti-Head AttentionResidual ConnectionSoftmaxWeight Decay

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections