Papers › Revisiting Label Smoothing and Knowledge Distillation Compatibility: What was Missing?

Revisiting Label Smoothing and Knowledge Distillation Compatibility: What was Missing?

29 Jun 2022arXiv:2206.14532archive 2025-07-28

Keshigeyan Chandrasegaran, Ngoc-Trung Tran, Yunqing Zhao, Ngai-Man Cheung

This work investigates the compatibility between label smoothing (LS) and knowledge distillation (KD). Contemporary findings addressing this thesis statement take dichotomous standpoints: Muller et al. (2019) and Shen et al. (2021b). Critically, there is no effort to understand and resolve these contradictory findings, leaving the primal question -- to smooth or not to smooth a teacher network? -- unanswered. The main contributions of our work are the discovery, analysis and validation of systematic diffusion as the missing concept which is instrumental in understanding and resolving these contradictory findings. This systematic diffusion essentially curtails the benefits of distilling from an LS-trained teacher, thereby rendering KD at increased temperatures ineffective. Our discovery is comprehensively supported by large-scale experiments, analyses and case studies including image classification, neural machine translation and compact student distillation tasks spanning across multiple datasets and teacher-student architectures. Based on our analysis, we suggest practitioners to use an LS-trained teacher with a low-temperature transfer to achieve high performance students. Code and models are available at https://keshik6.github.io/revisiting-ls-kd-compatibility/

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 16 code samples harvested from 1 repository linked to this paper: 10 ran (of which 0 constructed an object rather than computing a result; 9 with no instrument failure: 0 honoured, 0 violated, 9 with no contract checked; 1 where Syntology's instrument failed) · 6 unverified. Where Syntology's instrument failed: 1 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 16 samples from 1 repository, 10 ran (9 with no instrument failure; 1 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 2026-09-24 and 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 1 of the 10 samples that ran has no recorded date.

sutd-visual-computing-group/LS-KD-compatibility officialmentioned on GitHubpytorchMITSyntology: 16 samples, 10 ran (1 instrument-failed) · 6 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

16 samples harvested: 10 ran (of which 0 constructed an object rather than computing a result; 9 with no instrument failure: 0 honoured, 0 violated, 9 with no contract checked; 1 where Syntology's instrument failed) · 6 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran; the contract check could not run
9ran
6unverified

Licence: 1 of the 16 samples is pointer only, meaning Syntology does not serve that copy's text. For that sample, 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 sutd-visual-computing-group/LS-KD-compatibility. “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.

sutd-visual-computing-group/LS-KD-compatibility official repository 16 samples, 10 ran (1 instrument-failed) · 6 unverified
accuracy in 872 papers
census: 8 implementations from 7 papers, 7 distinct outputs across 6 buckets, one shared input per bucket
src/image_classification/imagenet/official_evaluation.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant MIT (permissive) · 131a82fd65128218 · report fixture: signature output:Tensor/2/float, target:Tensor/1/int, topk:tuple/1/int · returned list · level recorded, date not recorded
adjust_learning_rate in 125 papers src/image_classification/cub/train_student.py ran MIT (permissive) · 4404239a85829756 · report fixture: signature optimizer:SGD/0/?, epoch:int/0/int, step:int/0/int, len_epoch:int/0/int, args:Args/0/? · returned float · level recorded 2026-09-25
adjust_learning_rate in 125 papers src/image_classification/cub/train_teacher.py ran MIT (permissive) · e93943edd5a8472f · report fixture: signature optimizer:SGD/0/?, epoch:int/0/int, step:int/0/int, len_epoch:int/0/int, args:Args/0/? · returned float · level recorded 2026-09-25
adjust_learning_rate in 125 papers src/image_classification/imagenet/official_evaluation.py ran MIT (permissive) · fa122dd6d0671258 · report fixture: signature optimizer:SGD/0/?, epoch:int/0/int, args:Args/0/? · returned float · level recorded 2026-09-25
get_imagenet_trained_teacher_for_imagenet_student src/image_classification/imagenet/utils.py ran MIT (permissive) · fbaa9853176fc776 · report fixture: signature args:Args/0/? · returned tuple · level recorded 2026-09-25
to_python_float in 10 papers src/image_classification/imagenet/train_student_dali.py ran MIT (permissive) here; the stored copy is pointer only · 6b5d363ab4dbf012 · report fixture: signature t:Tensor/0/float · returned float · level recorded 2026-09-24
train in 522 papers src/image_classification/cub/train_student.py ran MIT (permissive) · 2a0d7e917240db2a · report fixture: signature train_loader:list/1/tuple, model:callable/0/?, teacher:callable/0/?, criterion:callable/0/?, optimizer:SGD/0/?, scaler:GradScaler/0/?, epoch:int/0/int, args:Args/0/? · returned tuple · level recorded 2026-09-25
validate in 313 papers src/image_classification/cub/train_student.py ran MIT (permissive) · 020b5acc2cea526a · report fixture: signature val_loader:list/1/tuple, model:callable/0/?, criterion:callable/0/?, args:Args/0/? · returned tuple · level recorded 2026-09-25
validate in 313 papers src/image_classification/imagenet/official_evaluation.py ran MIT (permissive) · 889998947e765d45 · report fixture: signature val_loader:list/1/tuple, model:callable/0/?, criterion:callable/0/?, args:Args/0/? · returned tuple · level recorded 2026-09-25
validate in 313 papers src/image_classification/imagenet/.ipynb_checkpoints/official_evaluation-checkpoint.py ran MIT (permissive) · 00c8fdc297e7e272 · report fixture: signature val_loader:list/1/tuple, model:callable/0/?, criterion:callable/0/?, args:Args/0/? · returned tuple · level recorded 2026-09-25
get_cub_trained_teacher_for_cub_student src/image_classification/cub/utils.py unverified MIT (permissive) · 205d6e975fc4fca7 · report
get_imagenet_trained_teacher_for_cub_teacher src/image_classification/cub/utils.py unverified MIT (permissive) · 35f1a5ee27e2d2a0 · report
get_penultimate_fts src/visualization/alpha-LS-KD_imagenet_centroids.py unverified MIT (permissive) · 06ec1746d47a9901 · report
get_test_loader in 10 papers src/visualization/data_loader_imagenet.py unverified MIT (permissive) · afe574de7fe140de · report
load_model_kd src/visualization/alpha-LS-KD_imagenet_centroids.py unverified MIT (permissive) · 33ec9768e54f5c4a · report
load_model_ls src/visualization/alpha-LS-KD_imagenet_centroids.py unverified MIT (permissive) · 1517519d4d7d27cf · report

Tasks

Image ClassificationKnowledge DistillationMachine Translationimage-classification

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

DiffusionKnowledge DistillationLabel Smoothing

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