The paper introduces Semantic Head Specialization (SHS), a phenomenon where Vision Transformer (ViT) attention heads specialize as either object- or background-focused, most evident under full attention. It proposes the SHS-Index to quantify this specialization, demonstrating its ability to distinguish full-attention from chunk-window ViTs and its strong correlation with downstream benchmark performance. Leveraging insights into window interaction, token serialization, and local softmax allocation, the authors design Ariadne Attention, a hybrid attention mechanism that matches full-attention performance on 22 image and video tasks while reducing attention compute by 6.5×.
By Chenhong He, Lei Li, Shicheng Li, Hanglong Lv, Lingpeng Kong, Qi Liu, Tong Yang, Shuhuai Ren
arXiv:2410.14874v3 Announce Type: replace
Abstract: Multi-Head Self-Attention (MHSA) is the cornerstone of Vision Transformers, allowing models to capture diverse feature representations by projectin...
By Tianxiao Zhang, Bo Luo, Guanghui Wang
HiLRP introduces a unified attribution framework for Vision Transformers (ViTs) that addresses the challenges posed by diverse architectural designs. By decomposing ViT operations into four basic types—linear maps, bilinear mixing, normalization/gating, and reindexing—HiLRP applies conservation‑satisfying relevance rules, enabling reliable explanations across a wide range of backbones. The method outperforms 14 existing attribution techniques on 10 architectures, maintaining conservation and improving localization accuracy (0.97 Pointing) compared to competitors.
By Sathiyamohan Nishankar, Pubudu Sanjeewani, Asanka Perera, Selvarajah Thuseethan
arXiv:2605. 18848v3 Announce Type: replace Abstract: This paper introduces Exact Linear Attention (ELA), a mechanism that achieves linear computational complexity for Transformer attention by exploiting the exact decomposition property of kernel functions, thereby eliminating approximation error.
By Weinuo Ou
arXiv:2602. 06883v3 Announce Type: replace Abstract: The smoothness of the transformer architecture has been extensively studied in the context of generalization, training stability, and adversarial robustness.
By Ambroise Odonnat, Laetitia Chapel, Romain Tavenard, Ievgen Redko
arXiv:2608.06205v2 Announce Type: replace
Abstract: Multispectral object detection combines visible and thermal imagery to improve perception under challenging illumination and environmental conditio...
By Nima Hatami, Karim Faez, Saeed Sharifian, Hamidreza Amindavar
arXiv:2606. 27449v1 Announce Type: new Abstract: Multi-head attention conventionally partitions the hidden dimension equally across all heads at every layer, enforcing an identical representational subspace dimension (dh = dmodel/h) throughout the models depth.
By Shubham Aggarwal
The paper introduces an importance‑scoring metric for multi‑head transformer attention heads applied to tabular data, a domain where transformers have been less studied. Experiments on 40 diverse tabular datasets show that removing heads with the lowest importance scores has minimal impact on performance, while removing the most important head first causes the largest drop. The study finds that important heads are distributed across layers and vary significantly across different tabular schemas, suggesting that the proposed score can help reduce redundancy and improve transformer efficiency.
By Ahmad Jad Allah, Kazi F. Akhter, Md. Kamrozzaman Bhuiyan, Manar D. Samad
The paper introduces Cubit, a Transformer‑style architecture that replaces the standard attention mechanism with Kernel Ridge Regression (KRR). By interpreting attention as Nadaraya‑Watson regression, Cubit incorporates the closed‑form KRR solution, combining kernel‑based value aggregation with normalization via the inverse kernel matrix. The authors also propose a Limited‑Range Rescale (LRR) to stabilize training and report that Cubit shows improved long‑sequence modeling, with gains increasing as training sequence length grows.
By Chuanyang Zheng, Jiankai Sun, Yihang Gao, Yuehao Wang, Liangchen Tan, Mac Schwager, Anderson Schneider, Yuriy Nevmyvaka, Xiaodong Liu
arXiv:2603.26380v3 Announce Type: replace
Abstract: The attention mechanism has been the core component in modern transformer architectures. However, the computation of standard full attention scales...
By Yusheng Zhao, Hourun Li, Bohan Wu, Yichun Yin, Lifeng Shang, Jingyang Yuan, Meng Zhang, Ming Zhang
arXiv:2607. 11990v1 Announce Type: cross Abstract: Feedforward network (FFN) blocks account for a large fraction of the parameters and computation in Transformer architectures, yet their internal structure remains difficult to interpret due to the additive superposition induced by the residual stream.
By Johannes Knittel, Hanspeter Pfister
arXiv:2607. 03612v1 Announce Type: cross Abstract: Feed-forward 3D reconstruction (F3R) transformers have recently achieved remarkable success.
By Jianing Deng, Yuanzhe Li, Jialu Wang, Song Wang, Tianlong Chen, Huanrui Yang, Jingtong Hu