Papers › Fantastic Weights and How to Find Them: Where to Prune in Dynamic Sparse Training

Fantastic Weights and How to Find Them: Where to Prune in Dynamic Sparse Training

21 Jun 2023NeurIPS 2023 11arXiv:2306.12230archive 2025-07-28

Aleksandra I. Nowak, Bram Grooten, Decebal Constantin Mocanu, Jacek Tabor

Dynamic Sparse Training (DST) is a rapidly evolving area of research that seeks to optimize the sparse initialization of a neural network by adapting its topology during training. It has been shown that under specific conditions, DST is able to outperform dense models. The key components of this framework are the pruning and growing criteria, which are repeatedly applied during the training process to adjust the network's sparse connectivity. While the growing criterion's impact on DST performance is relatively well studied, the influence of the pruning criterion remains overlooked. To address this issue, we design and perform an extensive empirical analysis of various pruning criteria to better understand their impact on the dynamics of DST solutions. Surprisingly, we find that most of the studied methods yield similar results. The differences become more significant in the low-density regime, where the best performance is predominantly given by the simplest technique: magnitude-based pruning. The code is provided at https://github.com/alooow/fantastic_weights_paper

PaperPDFConference PDFCodeCode 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 20 code samples harvested from 2 repositories linked to this paper: 15 ran (of which 0 constructed an object rather than computing a result; 13 with no instrument failure: 0 honoured, 0 violated, 13 with no contract checked; 2 where Syntology's instrument failed) · 5 unverified. Where Syntology's instrument failed: 2 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 12 samples from 1 repository, 9 ran (7 with no instrument failure; 2 where Syntology's instrument failed); found in paper text by Syntology: 8 samples from 1 repository, 6 ran (6 with no instrument failure; 0 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.

alooow/fantastic_weights_paper officialmentioned in paperpytorchMITSyntology: 12 samples, 9 ran (2 instrument-failed) · 3 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

20 samples harvested: 15 ran (of which 0 constructed an object rather than computing a result; 13 with no instrument failure: 0 honoured, 0 violated, 13 with no contract checked; 2 where Syntology's instrument failed) · 5 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.

2ran · our draft was wrong
13ran
5unverified

Licence: 5 of the 20 samples are pointer only, meaning Syntology does not serve that copy's text. For 5 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.

alooow/fantastic_weights_paper official repository 12 samples, 9 ran (2 instrument-failed) · 3 unverified
fast_collate in 11 papers imagenet_main.py ran MIT (permissive) · 717c76ec787ddfe7 · report fixture: signature batch:list/1/tuple · returned tuple · level recorded 2026-09-26
get_optimizer in 228 papers imagenet_main.py ran MIT (permissive) · 410b25b6d0e7a081 · report fixture: signature parameters:list/1/tuple, fp16:bool/0/bool, lr:float/0/float, momentum:float/0/float, weight_decay:float/0/float, true_wd:bool/0/bool, nesterov:bool/0/bool, state:NoneType/0/?, static_loss_scale:float/0/float, dynamic_loss_scale:bool/0/bool, bn_weight_decay:bool/0/bool · returned SGD · level recorded 2026-09-26
get_train_loader in 12 papers imagenet_main.py ran MIT (permissive) · 04b0a36942f99acf · report fixture: signature data_path:str/0/str, batch_size:int/0/int, workers:int/0/int, _worker_init_fn:NoneType/0/? · returned DataLoader · level recorded 2026-09-26
initializations models/initializers.py ran MIT (permissive) · 39e82a7c002940de · report fixture: signature init_type:str/0/str, density:float/0/float · returned function · level recorded 2026-09-26
magnitude_redistribution ImageNet/funcs.py ran MIT (permissive) here; the stored copy is pointer only · 7fbf3f0c2be8ea47 · report fixture: signature masking:NoneType/0/?, name:str/0/str, weight:Tensor/2/float, mask:Tensor/2/bool · returned float · level recorded 2026-09-25
momentum_redistribution ImageNet/funcs.py ran MIT (permissive) here; the stored copy is pointer only · 85588244473b88f9 · report fixture: signature masking:MockMasking/0/?, name:str/0/str, weight:Tensor/2/float, mask:Tensor/2/float · returned float · level recorded 2026-09-25
nonzero_redistribution ImageNet/funcs.py ran MIT (permissive) here; the stored copy is pointer only · e67bbb994ab5bae5 · report fixture: signature masking:NoneType/0/?, name:str/0/str, weight:Tensor/2/float, mask:NoneType/0/? · returned int · level recorded 2026-09-25
snip_forward_conv2d ImageNet/snip.py ran · our draft was wrong MIT (permissive) here; the stored copy is pointer only · 12d59cf9ada8b41a · report fixture: signature self:callable/0/?, x:Tensor/4/float · returned Tensor · level recorded 2026-09-22
snip_forward_linear ImageNet/snip.py ran · our draft was wrong MIT (permissive) here; the stored copy is pointer only · 94eeaaa88bd57941 · report fixture: signature self:callable/0/?, x:Tensor/2/float · returned Tensor · level recorded 2026-09-22
SNIP ImageNet/snip.py unverified MIT (permissive) · 04768d98372daca0 · report
build_resnet in 5 papers ImageNet/resnet.py unverified MIT (permissive) · f069023ee573fe51 · report
build_resnet in 5 papers models/imagenet_resnet.py unverified MIT (permissive) · 61bbac980274b60d · report
TimDettmers/sparse_learning found in paper text by Syntology 8 samples, 6 ran (0 instrument-failed) · 2 unverified
diverging_colors plot_graphs.py ran MIT (permissive) · 20c4b599af53b4ed · report fixture: signature n:int/0/int · returned list · level recorded 2026-09-26
rainbow_colors plot_graphs.py ran MIT (permissive) · b8a31ec73b7a8b3c · report fixture: signature n:int/0/int · returned list · level recorded 2026-09-26
sequential_colors plot_graphs.py ran MIT (permissive) · 4746465c14058f8a · report fixture: signature n:int/0/int · returned list · level recorded 2026-09-26
uniform_coverage imagenet/baseline/parameterized_tensors.py ran MIT (permissive) · bb11b6868648aca9 · report fixture: signature rank:int/0/int, n_features:int/0/int · returned Tensor · level recorded 2026-09-26
variance_redistribution mnist_cifar/extensions.py ran MIT (permissive) · 177842ec69eed409 · report fixture: signature masking:Masking/0/?, name:str/0/str, weight:Tensor/2/float, mask:Tensor/2/bool · returned float · level recorded 2026-09-26
your_pruning mnist_cifar/extensions.py ran MIT (permissive) · 77b9016770359ac8 · report fixture: signature masking:Tensor/2/float, mask:Tensor/2/int, weight:Tensor/2/float, name:str/0/str · returned Tensor · level recorded 2026-09-26
build_resnet in 5 papers imagenet/tuned_resnet/resnet.py unverified MIT (permissive) · 51a6a5c2a049e900 · report
your_redistribution mnist_cifar/extensions.py unverified MIT (permissive) · 37199b84c6e58a0e · report

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

DSTPruning

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