Copertina

Colinearity Decay: Training Quantization-Friendly ViTs with Outlier Decay

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Tong, Jin, Liang, Guang, Sun, Peilin, Wu, Jianxin
Natura:	Preprint
Pubblicazione:	2026
Soggetti:	Computer Vision and Pattern Recognition
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Nonlinear Bipolar Compensation: Handling Outliers in Post-Training Quantization
di: Sun, Peilin, et al.
Pubblicazione: (2026)

DopQ-ViT: Towards Distribution-Friendly and Outlier-Aware Post-Training Quantization for Vision Transformers
di: Yang, Lianwei, et al.
Pubblicazione: (2024)

I&S-ViT: An Inclusive & Stable Method for Pushing the Limit of Post-Training ViTs Quantization
di: Zhong, Yunshan, et al.
Pubblicazione: (2023)

CLAMP-ViT: Contrastive Data-Free Learning for Adaptive Post-Training Quantization of ViTs
di: Ramachandran, Akshat, et al.
Pubblicazione: (2024)

ADFQ-ViT: Activation-Distribution-Friendly Post-Training Quantization for Vision Transformers
di: Jiang, Yanfeng, et al.
Pubblicazione: (2024)

U-REPA: Aligning Diffusion U-Nets to ViTs
di: Tian, Yuchuan, et al.
Pubblicazione: (2025)

Harnessing the Computation Redundancy in ViTs to Boost Adversarial Transferability
di: Liu, Jiani, et al.
Pubblicazione: (2025)

Purrturbed but Stable: Human-Cat Invariant Representations Across CNNs, ViTs and Self-Supervised ViTs
di: Shah, Arya, et al.
Pubblicazione: (2025)

Elastic ViTs from Pretrained Models without Retraining
di: Simoncini, Walter, et al.
Pubblicazione: (2025)

Pretrained ViTs Yield Versatile Representations For Medical Images
di: Matsoukas, Christos, et al.
Pubblicazione: (2023)

Training-Free Acceleration of ViTs with Delayed Spatial Merging
di: Heo, Jung Hwan, et al.
Pubblicazione: (2023)

STRAP-ViT: Segregated Tokens with Randomized -- Transformations for Defense against Adversarial Patches in ViTs
di: Chattopadhyay, Nandish, et al.
Pubblicazione: (2026)

UniRefiner: Teaching Pre-trained ViTs to Self-Dispose Dross via Contrastive Register
di: Qiu, Congpei, et al.
Pubblicazione: (2026)

MPTQ-ViT: Mixed-Precision Post-Training Quantization for Vision Transformer
di: Tai, Yu-Shan, et al.
Pubblicazione: (2024)

TWEO: Transformers Without Extreme Outliers Enables FP8 Training And Quantization For Dummies
di: Liang, Guang, et al.
Pubblicazione: (2025)

Intriguing Frequency Interpretation of Adversarial Robustness for CNNs and ViTs
di: Chen, Lu, et al.
Pubblicazione: (2025)

Unsupervised Object Localization in the Era of Self-Supervised ViTs: A Survey
di: Siméoni, Oriane, et al.
Pubblicazione: (2023)

Parameter Efficient Fine-tuning of Self-supervised ViTs without Catastrophic Forgetting
di: Bafghi, Reza Akbarian, et al.
Pubblicazione: (2024)

Token Cropr: Faster ViTs for Quite a Few Tasks
di: Bergner, Benjamin, et al.
Pubblicazione: (2024)

Octic Vision Transformers: Quicker ViTs Through Equivariance
di: Nordström, David, et al.
Pubblicazione: (2025)

EdgeCrafter: Compact ViTs for Edge Dense Prediction via Task-Specialized Distillation
di: Liu, Longfei, et al.
Pubblicazione: (2026)

AFIDAF: Alternating Fourier and Image Domain Adaptive Filters as an Efficient Alternative to Attention in ViTs
di: Zheng, Yunling, et al.
Pubblicazione: (2024)

Which Direction to Choose? An Analysis on the Representation Power of Self-Supervised ViTs in Downstream Tasks
di: Kaltampanidis, Yannis, et al.
Pubblicazione: (2025)

Adapting Pretrained ViTs with Convolution Injector for Visuo-Motor Control
di: Hwang, Dongyoon, et al.
Pubblicazione: (2024)

VIVID-Med: LLM-Supervised Structured Pretraining for Deployable Medical ViTs
di: Wang, Xiyao, et al.
Pubblicazione: (2026)

APHQ-ViT: Post-Training Quantization with Average Perturbation Hessian Based Reconstruction for Vision Transformers
di: Wu, Zhuguanyu, et al.
Pubblicazione: (2025)

TAP-ViTs: Task-Adaptive Pruning for On-Device Deployment of Vision Transformers
di: Wang, Zhibo, et al.
Pubblicazione: (2026)

ViTs for Action Classification in Videos: An Approach to Risky Tackle Detection in American Football Practice Videos
di: Zaidi, Syed Ahsan Masud, et al.
Pubblicazione: (2026)

CAIT: Triple-Win Compression towards High Accuracy, Fast Inference, and Favorable Transferability For ViTs
di: Wang, Ao, et al.
Pubblicazione: (2023)

ConcatPlexer: Additional Dim1 Batching for Faster ViTs
di: Han, Donghoon, et al.
Pubblicazione: (2023)

ViTs are Everywhere: A Comprehensive Study Showcasing Vision Transformers in Different Domain
di: Mia, Md Sohag, et al.
Pubblicazione: (2023)

Exploring the Synergies of Hybrid CNNs and ViTs Architectures for Computer Vision: A survey
di: Yunusa, Haruna, et al.
Pubblicazione: (2024)

Causality $\neq$ Decodability, and Vice Versa: Lessons from Interpreting Counting ViTs
di: Huang, Lianghuan, et al.
Pubblicazione: (2025)

DenseNets Reloaded: Paradigm Shift Beyond ResNets and ViTs
di: Kim, Donghyun, et al.
Pubblicazione: (2024)

Communication Efficient Split Learning of ViTs with Attention-based Double Compression
di: Alvetreti, Federico, et al.
Pubblicazione: (2025)

Register and [CLS] tokens yield a decoupling of local and global features in large ViTs
di: Lappe, Alexander, et al.
Pubblicazione: (2025)

DFQ-ViT: Data-Free Quantization for Vision Transformers without Fine-tuning
di: Tong, Yujia, et al.
Pubblicazione: (2025)

QID: Efficient Query-Informed ViTs in Data-Scarce Regimes for OCR-free Visual Document Understanding
di: Le, Binh M., et al.
Pubblicazione: (2025)

Decomposing and Interpreting Image Representations via Text in ViTs Beyond CLIP
di: Balasubramanian, Sriram, et al.
Pubblicazione: (2024)

ViT$^3$: Unlocking Test-Time Training in Vision
di: Han, Dongchen, et al.
Pubblicazione: (2025)