The paper investigates how attention dynamics evolve across recurrent depth in language models, finding that attention support stabilizes early while hidden states and outputs take longer. It proposes WISE, a training‑free method that uses full attention in early steps and then reuses the discovered sparse working set for later steps, preserving performance on multi‑hop QA tasks. Experiments show that WISE maintains quality up to 2K context, offers measurable speedups, and highlights the importance of recurrent discovery of attention support.
By Ke Wan, Chen Chen
The paper introduces NAMOH, a native sparse attention mechanism that activates only a subset of heads per token, allowing each head to attend to a limited subsequence of tokens. By scaling the number of heads while keeping the active heads per token fixed, the method shortens head histories and reduces key‑value access without increasing overall storage. Experiments demonstrate that NAMOH can outperform fully activated models with the same parameter count and enable more efficient long‑context inference than smaller dense models.
By Zizhuo Fu, Runsheng Wang, Meng Li
arXiv:2607. 19358v1 Announce Type: new Abstract: Recent advances in long chain-of-thought reasoning models such as DeepSeek-R1 have led to increasingly longer inference context lengths under the test-time scaling paradigm.
By Yu Zhao, Zekun Zhang, Fan Jiang, Bo Zeng, Linlong Xu, Shimin Shan, Yu Liu, Longyue Wang, Weihua Luo
arXiv:2609.07086v1 Announce Type: new
Abstract: Transformers are a dominant architecture in modern machine learning, powering applications across vision, language, and beyond. At the core of their su...
By Hemanth Saratchandran, Simon Lucey
The paper introduces Declarative Attention (DA), a protocol that lets language models explicitly declare which parts of their context to focus on during generation. By partitioning decoding into full-context, region-specific, and recent-output-only modes, the inference engine can skip large portions of the KV cache, dramatically reducing attended tokens. Experiments on 15 long-context tasks with off-the-shelf models show significant savings (52.0% and 31.1% reductions) with only modest accuracy drops that diminish as model size increases.
By Namgyu Ho, Huzama Ahmad, Woosung Koh, Se-Young Yun, Tal Schuster, Cicero Nogueira dos Santos
RAVE (Re-Allocating Visual Attention) is a lightweight pair‑gating mechanism that adds a learned query‑key bias to pre‑softmax attention scores over visual keys, derived from pre‑RoPE query and key features. It requires no architectural changes to the backbone and can be trained end‑to‑end with the rest of the model. Across multiple multimodal benchmarks, RAVE improves standard attention by an average of 3 points, especially on perception‑intensive tasks such as multilingual OCR, chart understanding, document VQA, and scene text VQA.
By Xi Leng, Xinhong Ma, Ziqiang Dong, Feng Zhang, Xiaoying Tang, Yang Yang, Guanjun Jiang
arXiv:2512.20636v2 Announce Type: replace-cross
Abstract: Many self-attention sublayers in large language models (LLMs) can be removed with little to no loss. We attribute this to the Attention Suppr...
By Dhananjay Saikumar, Blesson Varghese
arXiv:2505. 15548v2 Announce Type: replace Abstract: Autoregressive transformer language models frequently exhibit training instability when trained on long sequences, particularly under low-precision arithmetic.
By Suvadeep Hajra
arXiv:2608. 06776v1 Announce Type: new Abstract: A standard self-attention layer consists of two interacting circuits: the query-key circuit that governs attention allocation, and the output-value circuit that maps attended representations to predictions.
By Rahul Vashisht, Harish G. Ramaswamy
arXiv:2601. 11667v2 Announce Type: replace-cross Abstract: Transformer architectures deliver state-of-the-art accuracy via dense full-attention, but their quadratic time and memory complexity with respect to sequence length limits practical deployment.
By Xiaojie Xia, Huigang Zhang, Chaoliang Zhong, Jun Sun, Yusuke Oishi
arXiv:2609.13141v1 Announce Type: new
Abstract: Post-training attention sparsification reduces the quadratic cumulative attention cost of pretrained Transformers by selecting a small set of context u...
By Zhiwei Li, Lei Zhu, Hao Gu, Xiang Hu, Yan Wang, Haitao Mi, Sirui Han, Leo Liang, Zhijiang Guo
arXiv:2606. 15378v1 Announce Type: cross Abstract: Modern language models increasingly adopt hybrid architectures that combine full attention with efficient attention modules, such as sliding-window attention (SWA) and recurrent sequence mixers.
By Ziqing Qiao, Yinuo Xu, Chaojun Xiao, Zhou Su, Zihan Zhou, Yingfa Chen, Xiaoyue Xu, Xu Han, Zhiyuan Liu