Papers › Joint Token Pruning and Squeezing Towards More Aggressive Compression of Vision Transformers

Joint Token Pruning and Squeezing Towards More Aggressive Compression of Vision Transformers

21 Apr 2023CVPR 2023 1arXiv:2304.10716archive 2025-07-28

Siyuan Wei, Tianzhu Ye, Shen Zhang, Yao Tang, Jiajun Liang

Although vision transformers (ViTs) have shown promising results in various computer vision tasks recently, their high computational cost limits their practical applications. Previous approaches that prune redundant tokens have demonstrated a good trade-off between performance and computation costs. Nevertheless, errors caused by pruning strategies can lead to significant information loss. Our quantitative experiments reveal that the impact of pruned tokens on performance should be noticeable. To address this issue, we propose a novel joint Token Pruning & Squeezing module (TPS) for compressing vision transformers with higher efficiency. Firstly, TPS adopts pruning to get the reserved and pruned subsets. Secondly, TPS squeezes the information of pruned tokens into partial reserved tokens via the unidirectional nearest-neighbor matching and similarity-based fusing steps. Compared to state-of-the-art methods, our approach outperforms them under all token pruning intensities. Especially while shrinking DeiT-tiny&small computational budgets to 35%, it improves the accuracy by 1%-6% compared with baselines on ImageNet classification. The proposed method can accelerate the throughput of DeiT-small beyond DeiT-tiny, while its accuracy surpasses DeiT-tiny by 4.78%. Experiments on various transformers demonstrate the effectiveness of our method, while analysis experiments prove our higher robustness to the errors of the token pruning policy. Code is available at https://github.com/megvii-research/TPS-CVPR2023.

PaperPDFConference PDFCodeCode Syntology ran

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

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2304.10716")

Code

Syntology Ran 2 of 8 code samples harvested from 1 repository linked to this paper; 6 have no recorded run. Of those that ran: 1 ran · our draft was wrong; 1 ran with no contract checked.

By repository: official repository: 8 samples from 1 repository, 2 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

megvii-research/tps-cvpr2023 officialmentioned in papermentioned on GitHubpytorchApache-2.0 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

8 samples harvested; 2 ran; 0 honoured the contract we drafted; 6 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

1ran · our draft was wrong
1ran
6unverified

Licence: 0 of the 8 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 megvii-research/TPS-CVPR2023. “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. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (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 from a repository marked official are listed first. 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 four 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; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). 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.

batch_index_select megvii-research/TPS-CVPR2023/torch_codebase/models/dynamic_vit.py official repository ran · our draft was wrong Apache-2.0 (permissive) · 908619be92da9903 · report
rand_bbox megvii-research/TPS-CVPR2023/megengine_codebase/mixup.py official repository ran Apache-2.0 (permissive) · 3913c620f6926f42 · report
build_dataset megvii-research/TPS-CVPR2023/torch_codebase/datasets.py official repository unverified Apache-2.0 (permissive) · 0d5eb6177974e3b2 · report
build_transform megvii-research/TPS-CVPR2023/torch_codebase/datasets.py official repository unverified Apache-2.0 (permissive) · 3bc74137ab36aa79 · report
contain_bias_and_norm megvii-research/TPS-CVPR2023/megengine_codebase/utils.py official repository unverified Apache-2.0 (permissive) · 427f72651ff7cfa0 · report
get_loss_cls megvii-research/TPS-CVPR2023/torch_codebase/losses/loss.py official repository unverified Apache-2.0 (permissive) · 7026a4f946a4f7a3 · report
get_module_cls megvii-research/TPS-CVPR2023/torch_codebase/models/dynamic_vit.py official repository unverified Apache-2.0 (permissive) · c0d55edbcda649c3 · report
get_sim megvii-research/TPS-CVPR2023/torch_codebase/models/dynamic_vit.py official repository unverified Apache-2.0 (permissive) · b31af37bba5ce554 · report

Tasks

Efficient ViTs

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Efficient ViTs ImageNet-1K (With LV-ViT-S) dTPS GFLOPs 3.8 #15 of 19 Archive leaderboard report
Efficient ViTs ImageNet-1K (With LV-ViT-S) dTPS Top 1 Accuracy 82.6 #15 of 19 Archive leaderboard report
Efficient ViTs ImageNet-1K (With LV-ViT-S) eTPS GFLOPs 3.8 #17 of 19 Archive leaderboard report
Efficient ViTs ImageNet-1K (With LV-ViT-S) eTPS Top 1 Accuracy 82.5 #17 of 19 Archive leaderboard report
Efficient ViTs ImageNet-1K (with DeiT-S) dTPS GFLOPs 3.0 #2 of 41 Archive leaderboard report
Efficient ViTs ImageNet-1K (with DeiT-S) dTPS Top 1 Accuracy 80.1 #2 of 41 Archive leaderboard report
Efficient ViTs ImageNet-1K (with DeiT-S) eTPS GFLOPs 3.0 #14 of 41 Archive leaderboard report
Efficient ViTs ImageNet-1K (with DeiT-S) eTPS Top 1 Accuracy 79.7 #14 of 41 Archive leaderboard report
Efficient ViTs ImageNet-1K (with DeiT-T) dTPS GFLOPs 0.8 #1 of 22 Archive leaderboard report
Efficient ViTs ImageNet-1K (with DeiT-T) dTPS Top 1 Accuracy 72.9 #1 of 22 Archive leaderboard report
Efficient ViTs ImageNet-1K (with DeiT-T) eTPS GFLOPs 0.8 #5 of 22 Archive leaderboard report
Efficient ViTs ImageNet-1K (with DeiT-T) eTPS Top 1 Accuracy 72.3 #5 of 22 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

Pruning

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