Papers › Distilling the Knowledge in a Neural Network
Distilling the Knowledge in a Neural Network
Geoffrey Hinton, Oriol Vinyals, Jeff Dean
A very simple way to improve the performance of almost any machine learning algorithm is to train many different models on the same data and then to average their predictions. Unfortunately, making predictions using a whole ensemble of models is cumbersome and may be too computationally expensive to allow deployment to a large number of users, especially if the individual models are large neural nets. Caruana and his collaborators have shown that it is possible to compress the knowledge in an ensemble into a single model which is much easier to deploy and we develop this approach further using a different compression technique. We achieve some surprising results on MNIST and we show that we can significantly improve the acoustic model of a heavily used commercial system by distilling the knowledge in an ensemble of models into a single model. We also introduce a new type of ensemble composed of one or more full models and many specialist models which learn to distinguish fine-grained classes that the full models confuse. Unlike a mixture of experts, these specialist models can be trained rapidly and in parallel.
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):
get_harvested_code_for_paper(arxiv_id="1503.02531")get_code(code_sha256="…")with a sample'scode_sha256from the list belowget_citation_path(paper_1="1503.02531", paper_2="…")with another paper's arXiv id or titleget_concepts_for_paper(arxiv_id="1503.02531")
Code
Syntology 37 code samples: 36 from 14 repositories linked to this paper, 1 identical to code harvested from another paper's repository. Across all 37: 23 ran (of which 0 constructed an object rather than computing a result; 14 with no instrument failure: 6 honoured, 0 violated, 8 with no contract checked; 9 where Syntology's instrument failed) · 14 unverified. Where Syntology's instrument failed: 8 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: community (archive-listed): 36 samples from 14 repositories, 23 ran (14 with no instrument failure; 9 where Syntology's instrument failed); 1 identical to code first harvested elsewhere. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.
No sample is linked to this paper from 13 of its repositories that Syntology harvested for other papers:
- Adlik/model_optimizer, community (archive-listed): harvested for another paper, listed under the repository
- altndrr/persona, community (archive-listed): harvested for another paper, listed under the repository
- aquvitae/aquvitae, community (archive-listed): harvested for another paper, listed under the repository
- hunto/image_classification_sota, community (archive-listed): harvested for 2 other papers, listed under the repository
- kaung-htet-myat/Multi-teachers-Knowledge-Distillation, community (archive-listed): harvested for another paper, listed under the repository
- markliou/model_distillation, community (archive-listed): harvested for another paper, listed under the repository
- mkavim/finetune_bert, community (archive-listed): harvested for 2 other papers, listed under the repository
- varungumma/fairseq, community (archive-listed): harvested for another paper, listed under the repository
- wannabeOG/ExpertNet-Pytorch, community (archive-listed): harvested for another paper, listed under the repository
- zh320/realtime-semantic-segmentation-pytorch, community (archive-listed): harvested for 4 other papers, listed under the repository
- Deci-AI/super-gradients, community (archive-listed): harvested for 4 other papers, listed under the repository
- alibaba/EasyNLP, community (archive-listed): harvested for another paper, listed under the repository
- labmlai/annotated_deep_learning_paper_implementations, community (archive-listed): harvested for 15 other papers, listed under the repository
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.
64 repositories listed; official and paper-mentioned ones first.
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.
A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.
Code Syntology ran Syntology
37 samples harvested: 23 ran (of which 0 constructed an object rather than computing a result; 14 with no instrument failure: 6 honoured, 0 violated, 8 with no contract checked; 9 where Syntology's instrument failed) · 14 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.
Licence: 11 of the 37 samples are pointer only, meaning Syntology does not serve that copy's text. For 1 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 14 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. 1 sample is listed last, in a group without a repository. “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.
0bd196791b1b86e6 · report
fixture: signature model_output:tuple/1/Tensor · returned dict · level recorded 2026-09-22 95712f5d0bd66d34 · report
fixture: signature device:device/0/?, batch_data:tuple/1/Tensor · returned tuple · level recorded 2026-09-22 5b365f68a2be17aa · report
7ee88cf415894459 · report
fixture: signature teacher_model:callable/0/?, dataloader:list/1/tuple · returned list · level recorded 2026-09-22 8a5635108617215c · report
fixture: signature outputs:Tensor/2/float, labels:Tensor/1/int, teacher_outputs:Tensor/2/float, params:Params/0/? · returned Tensor · level recorded 2026-09-22 2cbf06e40dbf54ee · report
fixture: signature dataset:MockDataset/0/? · returned tuple · level recorded 2026-09-22 9c20329610d8c5e5 · report
fixture: signature input:Tensor/2/float, model:callable/0/?, repeat:int/0/int · returned tuple · level recorded 2026-09-22 acb7b29614f1a161 · report
fixture: signature model:callable/0/?, thre:float/0/float, CBL_idx:list/1/int, prune_idx:list/1/int · returned tuple · level recorded 2026-09-22 ec78bec91ac01396 · report
fixture: signature model:callable/0/?, sorted_bn:Tensor/1/float, percent:float/0/float · returned Tensor · level recorded 2026-09-22 77551a1423d80be3 · report
fixture: signature path:str/0/str · returned ndarray · level recorded 2026-09-26 00c0d849e940be32 · report
fixture: signature num1:list/1/int, num2:list/1/int · returned bool · level recorded 2026-09-26 671c7f056497a8b6 · report
fixture: signature epoch:int/0/int, n_epoch:int/0/int, phase:str/0/str, loss:float/0/float, acc:float/0/float · returned str · level recorded 2026-09-26 a8b17275da494ca9 · report
fixture: signature nums:list/1/float · returned list · level recorded 2026-09-26 240a91c16bc48dd5 · report
fixture: signature preds:Tensor/2/float, labels:Tensor/1/int, teacher_preds:Tensor/2/float, T:float/0/float, alpha:float/0/float · returned Tensor · level recorded 2026-09-26 9f00b79fbcd75706 · report
fixture: signature x:Tensor/1/float · returned tuple · level recorded 2026-09-26 census: 53 implementations from 63 papers, 16 distinct outputs across 8 buckets, one shared input per bucket this sample: bucket 1 of 8, arg 1: rank 2, kind float, dtype float32 distiller/loss.py ran · honoured contract fingerprinted MIT (permissive) here; the stored copy is pointer only ·
cb5cbea0e8219ff5 · report
fixture: signature x:ndarray/2/float · returned ndarray · level recorded 2026-09-22 37db7f404db2f6ab · report
fixture: signature nums1:list/1/int, nums2:list/1/int · returned list · level recorded 2026-09-26 bee766c6e15044c2 · report
fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-24 f14f766b7e289b16 · report
978081881d93f4be · report
9eb8ac7c03e9b11b · report
274528b19a06a729 · report
fixture: signature student_features:Tensor/2/float, teacher_features:Tensor/2/float, topk:int/0/int, alpha:float/0/float, beta:float/0/float, gamma:float/0/float · returned Tensor · level recorded 2026-09-22 7308ca9b35a3b212 · report
fixture: signature model:callable/0/?, val_loader:DataLoader/0/?, criterion:callable/0/?, device:str/0/str · returned tuple · level recorded 2026-09-22 62f639fec5cb08e5 · report
fixture: signature model:callable/0/?, train_loader:DataLoader/0/?, criterion:callable/0/?, optimizer:SGD/0/?, device:str/0/str · returned tuple · level recorded 2026-09-22 00a58257f6886584 · report
fixture: signature train_dataset:SimpleDataset/0/?, val_dataset:SimpleDataset/0/?, model:callable/0/?, epochs:int/0/int, batch_size:int/0/int, device:str/0/str, opt:Adam/0/?, criterion:callable/0/? · returned ndarray · level recorded 2026-09-22 b7e8ebfd9b3a5233 · report
0a330946ecfac712 · report
fixture: signature in_channels:int/0/int, out_channels:int/0/int, batch_norm:bool/0/bool · returned Sequential · level recorded 2026-09-22 c3041253233293ca · report
a427a6169de8f607 · report
fixture: signature outputs:Tensor/2/float, targets:Tensor/1/int, topk:tuple/1/int · returned list · level recorded 2026-09-22 1 sample is identical to code Syntology harvested from another paper's repository; this paper's own copy was not located, so no repository is named and no licence is recorded.
census: 8 implementations from 7 papers, 7 distinct outputs across 6 buckets, one shared input per bucket repository not located unverified licence of this copy not recorded ·
9b8289076669fe4f · report
Tasks
Results from the paper archive 2025-07-28
| Task | Dataset | Model | Metric | Value | Rank at snapshot | Leaderboard | Report |
|---|---|---|---|---|---|---|---|
| Knowledge Distillation | CIFAR-100 | resnet8x4 (T: resnet32x4 S: resnet8x4) | Top-1 Accuracy (%) | 73.33 | #20 of 27 | Archive leaderboard | report |
| Knowledge Distillation | CIFAR-100 | vgg8 (T:vgg13 S:vgg8) | Top-1 Accuracy (%) | 72.98 | #21 of 27 | Archive leaderboard | report |
| Knowledge Distillation | CIFAR-100 | resnet110 (T:resnet110 S:resnet20) | Top-1 Accuracy (%) | 70.67 | #27 of 27 | Archive leaderboard | report |
| Knowledge Distillation | ImageNet | ADLIK-MO(T: ResNet101 S: ResNet50) | CRD training setting | ✘ | #22 of 52 | Archive leaderboard | report |
| Knowledge Distillation | ImageNet | ADLIK-MO(T: ResNet101 S: ResNet50) | Top-1 accuracy % | 77.14 | #22 of 52 | Archive leaderboard | report |
| Knowledge Distillation | ImageNet | ADLIK-MO(T: ResNet101 S: ResNet50) | model size | 99M | #22 of 52 | Archive leaderboard | report |
| Knowledge Distillation | ImageNet | KD (T: ResNet-34 S:ResNet-18) | CRD training setting | ✓ | #50 of 52 | Archive leaderboard | report |
| Knowledge Distillation | ImageNet | KD (T: ResNet-34 S:ResNet-18) | Top-1 accuracy % | 70.66 | #50 of 52 | Archive leaderboard | report |
Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.
Methods
Introduced by this paper: Knowledge Distillation
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