Pith. sign in

REVIEW 8 cited by

A Survey on Transformer Compression

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2402.05964 v2 pith:6QP4KWFG submitted 2024-02-05 cs.LG cs.CLcs.CV

A Survey on Transformer Compression

classification cs.LG cs.CLcs.CV
keywords compressionmethodsmodelslanguagelargetransformervisionarchitecture
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Transformer plays a vital role in the realms of natural language processing (NLP) and computer vision (CV), specially for constructing large language models (LLM) and large vision models (LVM). Model compression methods reduce the memory and computational cost of Transformer, which is a necessary step to implement large language/vision models on practical devices. Given the unique architecture of Transformer, featuring alternative attention and feedforward neural network (FFN) modules, specific compression techniques are usually required. The efficiency of these compression methods is also paramount, as retraining large models on the entire training dataset is usually impractical. This survey provides a comprehensive review of recent compression methods, with a specific focus on their application to Transformer-based models. The compression methods are primarily categorized into pruning, quantization, knowledge distillation, and efficient architecture design (Mamba, RetNet, RWKV, etc.). In each category, we discuss compression methods for both language and vision tasks, highlighting common underlying principles. Finally, we delve into the relation between various compression methods, and discuss further directions in this domain.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 8 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Depth Adaptive Efficient Visual Autoregressive Modeling

    cs.CV 2026-04 unverdicted novelty 7.0

    DepthVAR adaptively allocates per-token computational depth in VAR models using a cyclic rotated scheduler and dynamic layer masking to achieve 2.3-3.1x inference speedup with minimal quality loss.

  2. Do Transformers Need Three Projections? Systematic Study of QKV Variants

    cs.LG 2026-06 conditional novelty 6.0

    Q-K=V projection sharing in transformers matches standard QKV performance with 50% KV cache reduction and combines with GQA/MQA for up to 96.9% reduction across vision and language tasks.

  3. D-Legion: A Scalable Many-Core Architecture for Accelerating Matrix Multiplication in Quantized LLMs

    cs.AR 2026-02 unverdicted novelty 6.0

    D-Legion proposes a scalable architecture of Legions containing adaptive-precision systolic array cores that accelerates quantized LLM matrix multiplications, delivering up to 8.2x lower latency and 3.8x higher memory...

  4. Rethinking 1-bit Optimization Leveraging Pre-trained Large Language Models

    cs.CL 2025-08 conditional novelty 6.0

    A progressive training scheme with binary-aware initialization and dual-scaling allows pre-trained LLMs to be converted to high-performance 1-bit models without training from scratch.

  5. Distributional Statistics Restore Training Data Auditability in One-step Distilled Diffusion Models

    cs.LG 2025-02 unverdicted novelty 6.0

    Kernel-based distributional discrepancy enables auditing of upstream training data in distilled one-step diffusion models by detecting preserved distributional alignment rather than per-instance memorization.

  6. Distill-then-Replace: Efficient Task-Specific Hybrid Attention Model Construction

    cs.LG 2026-01 conditional novelty 5.0

    Distill-then-Replace builds task-specific hybrid attention LLMs by distilling each full-attention block into a linear counterpart and greedily replacing layers under a validation-performance constraint.

  7. MixFrag: Fragility-Guided Mixed-Precision Post-Training Quantization for Vision Transformers

    cs.CV 2026-07 conditional novelty 4.5

    KL-isolation fragility plus MCKP bit allocation yields mixed-precision ViT PTQ that lags recent ImageNet PTQ but reports large COCO AP gains at MP3/MP3.

  8. A Survey on Efficient Inference for Large Language Models

    cs.CL 2024-04 accept novelty 3.0

    The paper surveys techniques to speed up and reduce the resource needs of LLM inference, organized by data-level, model-level, and system-level changes, with comparative experiments on representative methods.