Papers › Decoupled Training for Long-Tailed Classification With Stochastic Representations

Decoupled Training for Long-Tailed Classification With Stochastic Representations

19 Apr 2023arXiv:2304.09426archive 2025-07-28

Giung Nam, Sunguk Jang, Juho Lee

Decoupling representation learning and classifier learning has been shown to be effective in classification with long-tailed data. There are two main ingredients in constructing a decoupled learning scheme; 1) how to train the feature extractor for representation learning so that it provides generalizable representations and 2) how to re-train the classifier that constructs proper decision boundaries by handling class imbalances in long-tailed data. In this work, we first apply Stochastic Weight Averaging (SWA), an optimization technique for improving the generalization of deep neural networks, to obtain better generalizing feature extractors for long-tailed classification. We then propose a novel classifier re-training algorithm based on stochastic representation obtained from the SWA-Gaussian, a Gaussian perturbed SWA, and a self-distillation strategy that can harness the diverse stochastic representations based on uncertainty estimates to build more robust classifiers. Extensive experiments on CIFAR10/100-LT, ImageNet-LT, and iNaturalist-2018 benchmarks show that our proposed method improves upon previous methods both in terms of prediction accuracy and uncertainty estimation.

PaperPDFCode Syntology ranMethod as code

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.

Syntology holds a reference implementation of SWA-Gaussian, generated from the paper text (not the authors' code), at level 2 on Syntology's ladder: V2 “two implementations GENERATED from the paper's signature and spec agree numerically (1e-6) on fixture inputs” “neither of the two ever read the authors' own code, so V2 measures how unambiguously the paper specifies the method, not fidelity to the authors' implementation (measured 2026-09-11, commit 75eaa77: cross_checked_by is a second LLM on 2,628 of 2,831 served samples)” (quoted from Syntology's server card). No rung compares this code against the authors' implementation. For agents: get_reference_implementation(name='SWA-Gaussian') (how to connect); the free check have('2304.09426') needs no token.

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

Code

Syntology 14 code samples harvested from 2 repositories linked to this paper: 10 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 0 honoured, 0 violated, 6 with no contract checked; 4 where Syntology's instrument failed) · 4 unverified. Where Syntology's instrument failed: 3 where our contract draft was wrong; 1 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: found in paper text by Syntology: 14 samples from 2 repositories, 10 ran (6 with no instrument failure; 4 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 between 2026-09-22 and 2026-09-26 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 2 of the 10 samples that ran have no recorded date.

cs-giung/long-tailed-srepr found in paper text by Syntologynot reachable when probed 2026-09-17 — repositories for recent papers often appear after camera-readySyntology: not harvested report
zhmiao/OpenLongTailRecognition-OLTR found in paper text by SyntologyBSD-3-ClauseSyntology: 9 samples, 6 ran (1 instrument-failed) · 3 unverified report
visipedia/inat found in paper text by Syntologynot reachable when probed 2026-09-17 — repositories for recent papers often appear after camera-readySyntology: not harvested report
kaidic/LDAM-DRW found in paper text by SyntologyMITSyntology: 5 samples, 4 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

14 samples harvested: 10 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 0 honoured, 0 violated, 6 with no contract checked; 4 where Syntology's instrument failed) · 4 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
1ran; the contract check could not run
6ran
4unverified

Licence: 2 of the 14 samples are pointer only, meaning Syntology does not serve that copy's text. For 2 of them, this paper's copy is under a permissive licence and cleared, but the stored copy of the same code that Syntology serves is not; the licence label's tooltip gives Syntology's recorded reason. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 2 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. “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. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

zhmiao/OpenLongTailRecognition-OLTR found in paper text by Syntology 9 samples, 6 ran (1 instrument-failed) · 3 unverified
conv3x3 in 1,376 papers models/ResNetFeature.py ran · our draft was wrong BSD-3-Clause (permissive) here; the stored copy is pointer only · fac5364e2f53c6db · report fixture: signature in_planes:int/0/int, out_planes:int/0/int, stride:int/0/int · returned Conv2d · level recorded, date not recorded
create_loss in 11 papers loss/DiscCentroidsLoss.py ran BSD-3-Clause (permissive) · 84954cc67863b073 · report fixture: signature feat_dim:int/0/int, num_classes:int/0/int · returned DiscCentroidsLoss · level recorded 2026-09-25
create_model in 163 papers models/CosNormClassifier.py ran BSD-3-Clause (permissive) · a50d9d62804228f2 · report fixture: signature in_dims:int/0/int, out_dims:int/0/int · returned CosNorm_Classifier · level recorded 2026-09-25
create_model in 163 papers models/DotProductClassifier.py ran BSD-3-Clause (permissive) · a2ab1cdb1975f18e · report fixture: signature feat_dim:int/0/int, num_classes:int/0/int, stage1_weights:bool/0/bool, dataset:NoneType/0/?, test:bool/0/bool · returned DotProduct_Classifier · level recorded 2026-09-25
shot_acc in 8 papers utils.py ran BSD-3-Clause (permissive) · ebb474138acbc60c · report fixture: signature preds:Tensor/1/int, labels:Tensor/1/int, train_data:DummyDataLoader/0/?, many_shot_thr:int/0/int, low_shot_thr:int/0/int · returned tuple · level recorded 2026-09-25
source_import in 5 papers utils.py ran BSD-3-Clause (permissive) · 07cca9631fe4f26a · report fixture: signature file_path:str/0/str · returned module · level recorded 2026-09-26
create_model in 163 papers models/ResNet10Feature.py unverified BSD-3-Clause (permissive) · aec4f3ee8eb5c0e4 · report
create_model in 163 papers models/ResNet152Feature.py unverified BSD-3-Clause (permissive) · 9b5c18aaeaac34a0 · report
init_weights in 109 papers utils.py unverified BSD-3-Clause (permissive) · 4826cd3596c12660 · report
kaidic/LDAM-DRW found in paper text by Syntology 5 samples, 4 ran (3 instrument-failed) · 1 unverified
accuracy in 872 papers
census: 8 implementations from 7 papers, 7 distinct outputs across 6 buckets, one shared input per bucket
utils.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant MIT (permissive) · 1e369e39c90cada6 · report fixture: signature output:Tensor/2/float, target:Tensor/1/int, topk:tuple/1/int · returned list · level recorded 2026-09-22
focal_loss in 65 papers
census: 6 implementations from 13 papers, 4 distinct outputs across 3 buckets, one shared input per bucket this sample: bucket 2 of 3, arg 1: rank 2, kind float, dtype float32
losses.py ran · our draft was wrong fingerprinted MIT (permissive) here; the stored copy is pointer only · 4dcf06ba28983266 · report fixture: signature input_values:Tensor/2/float, gamma:float/0/float · returned Tensor · level recorded, date not recorded
plot_confusion_matrix in 11 papers utils.py ran MIT (permissive) · e5ce72a1e8d9bc22 · report fixture: signature y_true:ndarray/1/int, y_pred:ndarray/1/int, classes:ndarray/1/<U7, normalize:bool/0/bool, title:str/0/str, cmap:callable/0/? · returned Axes · level recorded 2026-09-25
resnet32 in 47 papers models/resnet_cifar.py ran · our draft was wrong MIT (permissive) · 11d038f53d77a8d2 · report fixture: signature num_classes:int/0/int, use_norm:bool/0/bool · returned ResNet_s · level recorded 2026-09-25
validate in 313 papers cifar_train.py unverified MIT (permissive) · d9855945f564e8f2 · report

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. For a generated assembly, an independent peer is a second implementation generated separately from the same paper text; neither implementation read the authors' code. At level 2 the two agreed numerically on the fixture's inputs; level 3 is recorded when property tests derived from the paper also passed on both, and each level 3 card prints how many its report lists. A generated assembly's label was set when its code was generated for the method, and no model judged it; its card prints it as "by construction" or "not judged by a model". Where an assembly's phrase has no constructed figure, a brick that ran is not among the harvested samples this build reads, and whether a brick's run only constructed an object is read from that brick's harvested sample, so the figure was not counted; it does not say there were none.

1 assembly on this paper: 0 harvested, 1 generated.

generated Assembly for SWA-Gaussian (graph method: SWA-Gaussian), generated from the paper text, not the repository's code. 1 brick; individually on a fixture: 1 ran (1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified. Ran end-to-end against an independent peer (level 2).

label matches by construction: the code was generated for this method (not judged by a model)

  1. swag_diagonal_covariance entry · level 2 · ran · ceb587f090e330b2943191be55cd49b32121887ba359dae6d478e39d65b64f2c

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2304.09426") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

ClassificationRepresentation Learning

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

Stochastic Weight Averaging

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