Papers › Sparse autoencoders reveal selective remapping of visual concepts during adaptation

Sparse autoencoders reveal selective remapping of visual concepts during adaptation

6 Dec 2024arXiv:2412.05276archive 2025-07-28

Hyesu Lim, Jinho Choi, Jaegul Choo, Steffen Schneider

Adapting foundation models for specific purposes has become a standard approach to build machine learning systems for downstream applications. Yet, it is an open question which mechanisms take place during adaptation. Here we develop a new Sparse Autoencoder (SAE) for the CLIP vision transformer, named PatchSAE, to extract interpretable concepts at granular levels (e.g. shape, color, or semantics of an object) and their patch-wise spatial attributions. We explore how these concepts influence the model output in downstream image classification tasks and investigate how recent state-of-the-art prompt-based adaptation techniques change the association of model inputs to these concepts. While activations of concepts slightly change between adapted and non-adapted models, we find that the majority of gains on common adaptation tasks can be explained with the existing concepts already present in the non-adapted foundation model. This work provides a concrete framework to train and use SAEs for Vision Transformers and provides insights into explaining adaptation mechanisms.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

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

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 6 code samples harvested from 1 repository linked to this paper: 5 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 3 where Syntology's instrument failed) · 1 unverified. Where Syntology's instrument failed: 3 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 6 samples from 1 repository, 5 ran (2 with no instrument failure; 3 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written on 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

dynamical-inference/patchsae officialmentioned in papermentioned on GitHubjaxMITSyntology: 6 samples, 5 ran (3 instrument-failed) · 1 unverified 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

6 samples harvested: 5 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 3 where Syntology's instrument failed) · 1 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

3ran · our draft was wrong
2ran
1unverified

Licence: 0 of the 6 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 dynamical-inference/patchsae. “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. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (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 are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. 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 six 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; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. 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.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity.

dynamical-inference/patchsae official repository 6 samples, 5 ran (3 instrument-failed) · 1 unverified
calculate_entropy in 22 papers analysis/utils.py ran MIT (permissive) · c0113839c1642005 · report fixture: signature top_val:Tensor/2/float, top_label:Tensor/2/int, ignore_label_idx:NoneType/0/?, eps:float/0/float · returned Tensor · level recorded 2026-09-25
get_predictions in 17 papers tasks/classification_with_top_k_masking.py ran MIT (permissive) · 8b00d270f0a5a818 · report fixture: signature vit:callable/0/?, inputs:dict/0/?, text_features:Tensor/2/float, vit_type:str/0/str, hooks:NoneType/0/? · returned list · level recorded 2026-09-25
get_scheduler in 122 papers src/sae_training/utils.py ran · our draft was wrong MIT (permissive) · 4a7cb61e02537776 · report fixture: signature scheduler_name:str/0/str, optimizer:SGD/0/?, warm_up_steps:int/0/int, training_steps:int/0/int, lr_end:float/0/float, num_cycles:int/0/int · returned LambdaLR · level recorded 2026-09-25
load_checkpoint in 277 papers src/models/utils.py ran · our draft was wrong MIT (permissive) · b251e3420ace47d7 · report fixture: signature fpath:str/0/str · returned dict · level recorded 2026-09-25
load_state_dict_without_prompt_learner src/models/utils.py ran · our draft was wrong MIT (permissive) · effece23664132b2 · report fixture: signature ckpt_path:str/0/str · returned dict · level recorded 2026-09-25
load_stats in 5 papers analysis/utils.py unverified MIT (permissive) · daa2ff5af7df15ac · report

Tasks

Image Classificationimage-classification

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

CLIPSparse Autoencoder

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