Attention Mechanisms

Follow this topic Ask about this topic

Momentum

42 papers in the last four weeks, up 100% on the four weeks before. 0.4% of all new papers.

Jul 13Week of Sep 28

Latest papers 378

Apr 27, 2026cs.AR

Salca: A Sparsity-Aware Hardware Accelerator for Efficient Long-Context Attention Decoding

Long contexts improve capabilities of large language models but pose serious hardware challenges: compute and memory footprints grow linearly with sequence length. Particularly, the decoding phase continuously accesses massive KV cache, dramatically increasing bandwidth and computing pressure. Existing accelerators are primarily designed and evaluated for short contexts. They suffer from significant performance degradation when processing long contexts. To bridge this gap, we identify the major bottleneck and present a hardware accelerator for long context attention decoding via hardware-software co-design. On the software side, we propose dual-compression dynamic sparse attention. It combines ultra-low-precision quantization with feature sparsity to minimize prediction overhead. A hardware-friendly approximate Top-K selection further reduces filter complexity from

O(n \log k)

to

O(n)

. On the hardware side, we deeply optimize compute and memory access to tackle bottlenecks from intricate interplay between sparse attention and long contexts, and establish a performance model to derive the optimal co-design scheme. The resulting hardware adopts a fully pipelined parallel architecture and achieves

O(n)

efficiency even for long sequences. Experiments show that our design delivers

3.82\times

speedup and

74.19\times

energy efficiency over A100. Compared to SOTA accelerators, this is the first ASIC accelerator that efficiently supports long context inference, with at least

3.5\times

higher throughput and

2.08\times

better energy efficiency.

Wang Fan, Wei Cao, Xi Zha +5

Fudan University

Self-Attention Long-Context Language Model Inference LLM Inference Acceleration Sparse Attention Attention Mechanisms

Apr 27, 2026cs.CL

Kwai Summary Attention Technical Report

Long-context ability, has become one of the most important iteration direction of next-generation Large Language Models, particularly in semantic understanding/reasoning, code agentic intelligence and recommendation system. However, the standard softmax attention exhibits quadratic time complexity with respect to sequence length. As the sequence length increases, this incurs substantial overhead in long-context settings, leading the training and inference costs of extremely long sequences deteriorate rapidly. Existing solutions mitigate this issue through two technique routings: i) Reducing the KV cache per layer, such as from the head-level compression GQA, and the embedding dimension-level compression MLA, but the KV cache remains linearly dependent on the sequence length at a 1:1 ratio. ii) Interleaving with KV Cache friendly architecture, such as local attention SWA, linear kernel GDN, but often involve trade-offs among KV Cache and long-context modeling effectiveness. Besides the two technique routings, we argue that there exists an intermediate path not well explored: {Maintaining a linear relationship between the KV cache and sequence length, but performing semantic-level compression through a specific ratio

k

}. This

O(n/k)

path does not pursue a ``minimum KV cache'', but rather trades acceptable memory costs for complete, referential, and interpretable retention of long distant dependency. Motivated by this, we propose Kwai Summary Attention (KSA), a novel attention mechanism that reduces sequence modeling cost by compressing historical contexts into learnable summary tokens.

Chenglong Chu, Guorui Zhou, Guowang Zhang +35

Kuaishou

Self-Attention KV Caching Long-Context Language Model Inference Long-Context Language Modeling Transformer Attention Attention Mechanisms+1

Apr 27, 2026cs.LG

Nautile-370M: Spectral Memory Meets Attention in a Small Reasoning Model

We present Nautile-370M, a 371-million-parameter small language model designed for efficient reasoning under strict parameter and inference budgets. Nautile-370M uses a hybrid backbone in which two SeqCond Attention (SCA) layers, a linear-time spectral sequence operator inspired by SeqCondenser, alternate with one transformer layer. This design aims to retain the long-context efficiency and state-tracking benefits of structured sequential models while preserving the expressive token-to-token routing of attention. The model was trained on a single Cloud TPU v4-64 pod slice provided through the Google TPU Research Cloud (TRC) program; the subsequent reinforcement learning stage was carried out on a single NVIDIA DGX Spark. We prove that the SCA readout mechanism can exactly retrieve any individual token from the prefix summary and can reproduce any output of softmax attention as a special case, establishing that SCA is at least as expressive as full self-attention in the continuous limit. We also describe the training data pipeline and outline a reinforcement learning stage specialized for reasoning, verification, and response quality.

Maixent Chenebaux

Trickstr

Self-Attention Small Language Models Attention Mechanisms Efficient Language Model Inference Neural Network Expressivity

Apr 27, 2026cs.CL

Pruning via Causal Attribution Preserves Reasoning Performance in Large Language Models

Large language models (LLMs) excel at multi-step reasoning but incur substantial inference cost. We introduce Causal Attribution Pruning (CAP), a training-free method that identifies critical attention heads by measuring their causal impact on reasoning tasks and uses these head-level scores to guide fine-grained weight pruning. For each attention head, CAP estimates the expected performance degradation when the head is masked during forward passes on a small calibration set of reasoning problems. These causal scores are then converted into weight-level importance values for the corresponding projection matrices. Unlike magnitude-only or activation-based criteria, CAP's interventional measurement directly captures each head's functional contribution, yielding relative accuracy gains of up to 61% over Wanda on ARC-Challenge at 20% sparsity. We evaluate CAP on GSM8K, StrategyQA, and ARC-Challenge using Llama-3-8B-Instruct and Mistral-7B-Instruct at 10%, 20%, and 50% sparsity. At moderate sparsity (10-20%), CAP improves over Wanda in most model-benchmark configurations. with especially large gains on ARC-Challenge for Llama-3. Our results suggest that attention-head-level causal attribution can better preserve reasoning performance on downstream benchmarks than correlational pruning criteria at equivalent sparsity, while remaining limited by coarse MLP attribution at 50% sparsity.

Amogh Sheth, Biruk Assefa, Yi Wen Huang +2

Edison Academy Magnet School · State University of New York College at Plattsburgh · The University of Texas at Austin +2

LLM Pruning Self-Attention LLM Inference Acceleration Attention Mechanisms Attention Head Pruning Causal Interventions in Language Models+2

Apr 26, 2026cs.LG

ELSA: Exact Linear-Scan Attention for Fast and Memory-Light Vision Transformers

Existing attention accelerators often trade exact softmax semantics, depend on fused Tensor Core kernels, or incur sequential depth that limits FP32 throughput on long sequences. We present \textbf{ELSA}, an algorithmic reformulation of online softmax attention that (i)~preserves exact softmax semantics in real arithmetic with a \emph{provable}

\mathcal{O}(u\log n)

FP32 relative error bound; (ii)~casts the online softmax update as a prefix scan over an associative monoid

(m,S,W)

, yielding

O(n)

extra memory and

O(\log n)

parallel depth; and (iii)~is Tensor-Core independent, implemented in Triton and CUDA C++, and deployable as a \emph{drop-in replacement} requiring no retraining or weight modification. Unlike FlashAttention-2/3, which rely on HMMA/GMMA Tensor Core instructions and provide no compatible FP32 path, ELSA operates identically on A100s and resource-constrained edge devices such as Jetson TX2 -- making it the only hardware-agnostic exact-attention kernel that reduces parallel depth to

O(\log n)

at full precision. On A100 FP32 benchmarks (1K--16K tokens), ELSA delivers

1.3

--

3.5\times

speedup over memory-efficient SDPA and

1.97

--

2.27\times

on BERT; on Jetson TX2, ELSA achieves

1.5

--

1.6\times

over Math (64--900 tokens), with

17.8

--

20.2\%

throughput gains under LLaMA-13B offloading at

\ge

32K. In FP16, ELSA approaches hardware-fused baselines at long sequences while retaining full FP32 capability, offering a unified kernel for high-precision inference across platforms. Our code and implementation are available at https://github.com/ming053l/ELSA.

Chih-Chung Hsu, Xin-Di Ma, Wo-Ting Liao +1

Advanced Computer Vision Laboratory, National Yang Ming Chiao Tung University

Efficient Transformer Inference Softmax Attention GPU Acceleration Self-Attention High-Performance Computing Vision Transformer+3

Apr 26, 2026cs.CL

DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence

We present a preview version of DeepSeek-V4 series, including two strong Mixture-of-Experts (MoE) language models -- DeepSeek-V4-Pro with 1.6T parameters (49B activated) and DeepSeek-V4-Flash with 284B parameters (13B activated) -- both supporting a context length of one million tokens. DeepSeek-V4 series incorporate several key upgrades in architecture and optimization: (1) a hybrid attention architecture that combines Compressed Sparse Attention (CSA) and Heavily Compressed Attention (HCA) to improve long-context efficiency; (2) Manifold-Constrained Hyper-Connections (mHC) that enhance conventional residual connections; (3) and the Muon optimizer for faster convergence and greater training stability. We pre-train both models on more than 32T diverse and high-quality tokens, followed by a comprehensive post-training pipeline that unlocks and further enhances their capabilities. DeepSeek-V4-Pro-Max, the maximum reasoning effort mode of DeepSeek-V4-Pro, redefines the state-of-the-art for open models, outperforming its predecessors in core tasks. Meanwhile, DeepSeek-V4 series are highly efficient in long-context scenarios. In the one-million-token context setting, DeepSeek-V4-Pro requires only 27% of single-token inference FLOPs and 10% of KV cache compared with DeepSeek-V3.2. This enables us to routinely support one-million-token contexts, thereby making long-horizon tasks and further test-time scaling more feasible. The model checkpoints are available at https://huggingface.co/collections/deepseek-ai/deepseek-v4.

DeepSeek-AI, Anyi Xu, Bangcai Lin +316

Mixture-of-Experts Language Models Long-Context Language Model Inference Long-Context Language Modeling LLM Inference Acceleration Sparse Attention Attention Mechanisms+1

Apr 26, 2026cs.LG

Transformer as an Euler Discretization of Score-based Variational Flow

Despite the Transformer's dominance across machine learning, its architecture remains largely heuristic and lacks a unified theoretical foundation. We introduce Score-based Variational Flow (SVFlow), a continuous-time dynamical system for representation learning in which the state evolves according to a variational posterior-weighted average of conditional log-likelihood scores, and provide a principled basis for regularization through variational consistency. We show that forward Euler discretization of spherical SVFlow exactly recovers the Transformer architecture. Multi-head attention approximates SVFlow vector field via a vMF kernel-smoothed posterior, while MoE/FFN approximates it in a relaxed network-based way, and the residual-normalization block implements a relaxed retraction that maintains spherical geometry. This unification explains why attention trains stably without explicit regularization while MoE requires auxiliary balancing losses. Experiments on pre-trained language models with prefix shuffling show that SVFlow-induced metrics correlate with task performance, reveal depth-dependent sensitivity, and reflect the intrinsic dynamics of attention.

Huadong Liao

Representation Learning Transformer Self-Attention Decoder-Only Language Models Transformer Attention Attention Mechanisms

Apr 26, 2026cs.SD

HeadRouter: Dynamic Head-Weight Routing for Task-Adaptive Audio Token Pruning in Large Audio Language Models

Recent large audio language models (LALMs) demonstrate remarkable capabilities in processing extended multi-modal sequences, yet incur high inference costs. Token compression is an effective method that directly reduces redundant tokens in the sequence. Existing compression methods usually assume that all attention heads in LALMs contribute equally to various audio tasks and calculate token importance by averaging scores across all heads. However, our analysis demonstrates that attention heads exhibit distinct behaviors across diverse audio domains. We further reveal that only a sparse subset of attention heads actively responds to audio, with completely different performance when handling semantic and acoustic tasks. In light of this observation, we propose HeadRouter, a head-importance-aware token pruning method that perceives the varying importance of attention heads in different audio tasks to maximize the retention of crucial tokens. HeadRouter is training-free and can be applied to various LALMs. Extensive experiments on the AudioMarathon and MMAU-Pro benchmarks demonstrate that HeadRouter achieves state-of-the-art compression performance, exceeding the baseline model even when retaining 70% of the audio tokens and achieving 101.8% and 103.0% of the vanilla average on Qwen2.5-Omni-3B and Qwen2.5-Omni-7B, respectively.

Peize He, Yaodi Luo, Xiaoqian Liu +7

1EPIC Lab, Shanghai Jiao Tong University · 2DAIL Tech · 3Northeastern University +2

Audio Token Compression Self-Attention Audio Understanding Audio-Language Models Token Pruning Multimodal Token Compression+3

Apr 26, 2026cs.LG

Rank, Head-Channel Non-Identifiability, and Symmetry Breaking: A Precise Analysis of Representational Collapse in Transformers

A widely cited result by Dong et al. (2021) showed that Transformers built from self-attention alone, without skip connections or feed-forward layers, suffer from rapid rank collapse: all token representations converge to a single direction. The proposed remedy was the MLP. We show that this picture, while correct in the regime studied by Dong, is incomplete in ways that matter for architectural understanding. Three results are established. First, layer normalisation is precisely affine-rank-neutral: it preserves the affine rank of the token representation set exactly. The widespread claim that LN "plays no role" is imprecise; the correct statement is sharper. Second, residual connections generically obstruct rank collapse in real Transformers such as BERT-base, in a measure-theoretic sense, without contribution from the MLP. The MLP's irreplaceable function is different: generating feature directions outside the linear span of the original token embeddings, which no stack of attention layers can produce. Third, a phenomenon distinct from rank collapse is identified: head-channel non-identifiability. After multi-head attention sums per-head outputs through the output projection, individual contributions cannot be canonically attributed to a specific head; n(H-1)d_k degrees of freedom per layer remain ambiguous when recovering a single head from the mixed signal. The MLP cannot remedy this because it acts on the post-summation signal. A constructive partial remedy is proposed: a position-gated output projection (PG-OP) at parameter overhead below 1.6% of the standard output projection. The four collapse phenomena identified in the literature -- rank collapse in depth, in width, head-channel non-identifiability, and entropy collapse -- are unified under a symmetry-breaking framework, each corresponding to a distinct symmetry of the Transformer's forward pass.

Giansalvo Cirrincione

aLaboratoire LTI, Université de Picardie Jules Verne, Chemin du Thil, 80025 Amiens, France

Transformer Interpretability Representation Collapse Parameter Identifiability Self-Attention Symmetry Breaking Transformer Attention+2

Apr 24, 2026cs.CV

Different Strokes for Different Folks: Writer Identification for Historical Arabic Manuscripts

Handwritten Arabic manuscripts preserve the Arab world's intellectual and cultural heritage, and writer identification supports provenance, authenticity verification, and historical analysis. Using the Muharaf dataset of historical Arabic manuscripts, we evaluate writer identification from individual line images and, to the best of our knowledge, provide the first baselines reported under both line-level and page-disjoint evaluation protocols. Since the dataset is only partially labeled for writer identification, we manually verified and expanded writer labels in the public portion from 6,858 (28.00%) to 21,249 lines (86.75%) out of 24,495 line images, correcting inconsistencies and removing non-handwritten text. After further filtering, we retained 18,987 lines (77.51%). We propose a Convolutional Neural Network (CNN)-based model with attention mechanisms for closed-set writer identification, including rare two-writer lines modeled as composite writer-pair classes. We benchmark fourteen configurations and conduct ablations across different feature extractors and training regimes. To assess generalization to unseen pages, the page-disjoint protocol assigns all lines from each page to a single split. Under the line-level protocol, a fine-tuned DenseNet201 with attention achieves 99.05% Top-1 accuracy, 99.73% Top-5 accuracy, and 97.44% F1-score. Under the more challenging page-disjoint protocol, the best observed results are 78.61% Top-1 accuracy, 87.79% Top-5 accuracy, and 66.55% F1-score, thus quantifying the impact of page-level cues. By expanding the Muharaf dataset's labeled subset and reporting both protocols, we provide a clearer benchmark and a practical resource for historians and linguists engaged with culturally and historically significant documents. The code and implementation details are available on GitHub.

Hamza A. Abushahla, Ariel Justine N. Panopio, Layth Al-Khairulla +1

Department of Computer Science and Engineering, American University of Sharjah, Sharjah 26666, United Arab Emirates

Document Understanding Historical Document Analysis Authorship Attribution Attention Mechanisms Document Image Understanding Historical Document Processing

Apr 24, 2026cs.LG

HubRouter: A Pluggable Sub-Quadratic Routing Primitive for Hybrid Sequence Models

We introduce HubRouter, a pluggable module that replaces O(n^2) attention layers with O(nM) hub-mediated routing, where M << n is a small number of learned hub tokens. We demonstrate it in two from-scratch architectures: a Jamba-style hybrid and a 12-layer Transformer; retrofit into pretrained models is a tested negative case. HubRouter implements an encode-decode-score-council pipeline: M learned hubs cross-attend to all tokens, tokens project against hubs for routing fingerprints, a score head selects top-k tokens, and a sparse council attends only to the selected subset. We validate HubRouter in three settings. (1) Hub-Jamba yields a nominal 4.2% PPL improvement (200.2 vs 209.0, single seed; possibly within seed noise) and up to ~90x training throughput at sequence length 1024 in matched PyTorch-native baselines; an optimised baseline would narrow this to ~10-15x. (2) Graduated replacement of 25% of Transformer attention layers gives the best perplexity in our matched-budget sweep (268.0 vs 282.4 pure Transformer). (3) Hub-GPT provides strictly causal routing, achieving PPL 211.5 +/- 0.4 over 3 seeds (post council-causal fix); approximately 3 PPL worse than Jamba's 208.5 +/- 0.7, a measurable quality cost for avoiding O(n^2) computation. Post-fix, chunk size C has little effect; the pre-fix chunk-size benefit was an artifact of a bidirectional-council leak we found in adversarial review. A multi-seed hub-count sweep (~105 runs across M=1-32) reveals M=8-14 as the reliably-converging sub-band (4-5/5 seeds); M=6 is rescued to 5/5 by orthogonal regularization, while M>=20 shows increasing seed sensitivity. Companion paper arXiv:2603.20997 (Basu, 2026) defines the routing diagnostic task. Code and scripts will be released.

Abhinaba Basu

National Institute of Electronics and Information Technology (NIELIT)

Transformer Attention Language Modeling Sparse Attention Attention Mechanisms Sequence Modeling

Apr 24, 2026cs.CL

Dissociating Decodability and Causal Use in Bracket-Sequence Transformers

When trained on tasks requiring an understanding of hierarchical structure, transformers have been found to represent this hierarchy in distinct ways: in the geometry of the residual stream, and in stack-like attention patterns maintaining a last-in, first-out ordering. However, it remains unclear whether these representations are causally used or merely decodable. We examine this gap in transformers trained on the Dyck language (a formal language of balanced bracket sequences), where the hierarchical ground truth is explicit. By probing and intervening on the residual stream and attention patterns, we find that depth, distance, and top-of-stack signals are all decodable, yet their causal roles diverge. Specifically, masking attention to the true top-of-stack position causes a sharp drop in long-distance accuracy, while ablating low-dimensional residual stream subspaces has comparatively little effect. These results, which extend to a templated natural language setting, suggest that even in a controlled setting where the relevant hierarchical variables are known, decodability alone does not imply causal use.

Aryan Sharma, Cutter Dawes, Shivam Raval

1Yale University · 2Independent · 3Harvard University

Transformer Interpretability Self-Attention Transformer Attention Attention Mechanisms Causal Interpretability Causal Interventions in Language Models

Apr 23, 2026cs.CL

Shared Lexical Task Representations Explain Behavioral Variability In LLMs

One of the most common complaints about large language models (LLMs) is their prompt sensitivity -- that is, the fact that their ability to perform a task or provide a correct answer to a question can depend unpredictably on the way the question is posed. We investigate this variation by comparing two very different but commonly-used styles of prompting: instruction-based prompts, which describe the task in natural language, and example-based prompts, which provide in-context few-shot demonstration pairs to illustrate the task. We find that, despite large variation in performance as a function of the prompt, the model engages some common underlying mechanisms across different prompts of a task. Specifically, we identify task-specific attention heads whose outputs literally describe the task -- which we dub lexical task heads -- and show that these heads are shared across prompting styles and trigger subsequent answer production. We further find that behavioral variation between prompts can be explained by the degree to which these heads are activated, and that failures are at least sometimes due to competing task representations that dilute the signal of the target task. Our results together present an increasingly clear picture of how LLMs' internal representations can explain behavior that otherwise seems idiosyncratic to users and developers.

Zhuonan Yang, Jacob Xiaochen Li, Francisco Piedrahita Velez +5

Brown University · MIT · Northeastern

Prompt Sensitivity LLM Prompting LLM Interpretability Few-Shot Prompting Mechanistic Interpretability Attention Mechanisms

Apr 23, 2026cs.LG

Evaluating Post-hoc Explanations of the Transformer-based Genome Language Model DNABERT-2

Explaining deep neural network predictions on genome sequences enables biological insight and hypothesis generation-often of greater interest than predictive performance alone. While explanations of convolutional neural networks (CNNs) have been shown to capture relevant patterns in genome sequences, it is unclear whether this transfers to more expressive Transformer-based genome language models (gLMs). To answer this question, we adapt AttnLRP, an extension of layer-wise relevance propagation to the attention mechanism, and apply it to the state-of-the-art gLM DNABERT-2. Thereby, we propose strategies to transfer explanations from token and nucleotide level. We evaluate the adaption of AttnLRP on genomic datasets using multiple metrics. Further, we provide an extensive comparison between the explanations of DNABERT-2 and a baseline CNN. Our results demonstrate that AttnLRP yields reliable explanations corresponding to known biological patterns. Hence, like CNNs, gLMs can also help derive biological insights. This work contributes to the explainability of gLMs and addresses the comparability of relevance attributions across different architectures.

Isabel Kurth, Paulo Yanez Sarmiento, Bernhard Y. Renard

Hasso Plattner Institute University of Potsdam

Transformer Interpretability Attention Mechanisms Genomics Layer-Wise Relevance Propagation

Apr 23, 2026cs.LG

The Recurrent Transformer: Greater Effective Depth and Efficient Decoding

Transformers process tokens in parallel but are temporally shallow: at position

t

, each layer attends to key-value pairs computed based on the previous layer, yielding a depth capped by the number of layers. Recurrent models offer unbounded temporal depth but suffer from optimization instability and historically underutilize modern accelerators. We introduce the Recurrent Transformer, a simple architectural change where each layer attends to key-value pairs computed off its own activations, yielding layerwise recurrent memory while preserving standard autoregressive decoding cost. We show that the architecture can emulate both (i) a conventional Transformer and (ii) token-to-token recurrent updates under mild assumptions, while avoiding optimization instability. Naively, prefill/training appears bandwidth-bound with effective arithmetic intensity near

1

because keys and values are revealed sequentially; we give an exact tiling-based algorithm that preserves the mathematical computation while reducing HBM traffic from

Θ(N^2)

to

Θ(N\log N)

, increasing effective arithmetic intensity to

Θ(N/\log N)

for sequence length

N

. On 150M and 300M parameter C4 pretraining, Recurrent Transformers improve cross-entropy over a parameter-matched Transformer baseline and achieve the improvement with fewer layers (fixed parameters), suggesting that recurrence can trade depth for width, thus reducing KV cache memory footprint and inference latency.

Costin-Andrei Oncescu, Depen Morwani, Samy Jelassi +3

Harvard University

Efficient Transformer Inference Self-Attention Autoregressive Language Modeling Recurrent Transformers Efficient Attention Attention Mechanisms

Apr 22, 2026cs.CV

Attention-Augmented YOLOv8 with Ghost Convolution for Real-Time Vehicle Detection in Intelligent Transportation Systems

Accurate vehicle detection is a critical component of autonomous driving, traffic surveillance, and intelligent transportation systems. This paper presents an enhanced YOLOv8n-based model that integrates the Ghost Module, Convolutional Block Attention Module (CBAM), and Deformable Convolutional Networks v2 (DCNv2) to improve detection performance. The Ghost Module reduces feature redundancy through efficient feature generation, CBAM refines feature representation via channel and spatial attention, and DCNv2 enhances adaptability to geometric variations in vehicle structures. Evaluated on the KITTI dataset, the proposed model achieves 95.4% [email protected], representing an 8.97% improvement over the baseline YOLOv8n, along with 96.2% precision, 93.7% recall, and a 94.93% F1-score. Comparative analysis against seven state-of-the-art detectors demonstrates consistent superiority across key performance metrics, while ablation studies validate the individual and combined contributions of the integrated modules. By addressing feature redundancy, attention refinement, and spatial adaptability, the proposed approach offers a robust and computationally efficient solution for vehicle detection in diverse and complex traffic environments.

Syed Sajid Ullah, Muhammad Zunair Zamir, Ahsan Ishfaq +1

School of Energy and Electrical Engineering, Chang’an University, Xi’an, China · School of Information Engineering, Chang’an University, Xi’an, China

Intelligent Transportation Systems Object Detection Real-Time Object Detection Convolutional Neural Networks Attention Mechanisms Deformable Convolution

Apr 22, 2026cs.LG

Simplified Sparse Attention via Gist Tokens

Sparse attention can reduce the cost of long-context inference, but most variants introduce new architectural components. We introduce Simplified Sparse Attention (SSA), a simpler approach to sparse attention that requires no architectural changes. Concretely, we first perform continued pretraining on sequences interleaved with gist tokens. We optimize the standard next-token loss as usual, but the gist tokens use an attention mask to restrict what parts of the context the language model can attend to; this teaches the model to pack each chunk's important information into the gist tokens. At inference time, SSA scores chunks via attention between the current query and the small set of gist tokens, selectively unfolding the top-k chunks by reintroducing their corresponding raw tokens. Since the query is scored only against the gist tokens, we avoid the memory-bandwidth cost associated with naive scoring against the full KV cache, without requiring the auxiliary KV cache approach used by sparse attention methods. On LongBench, SSA consistently outperforms compression and inference-time sparse-attention baselines under the same compression ratio. More strikingly, in retrieval-augmented generation, SSA can even outperform full attention after continued pretraining by over 5.7 points. We attribute this to the ability of SSA's selective unfolding, which concentrates attention on the query-relevant chunks and effectively filters out noise. SSA further extends to a hierarchical gist-of-gist variant (H-SSA) that achieves log-linear decoding complexity while maintaining or improving accuracy at high compression ratios up to 32x. The code is available at https://github.com/yuzhenmao/simplified-sparse-attention/.

Yuzhen Mao, Michael Y. Li, Emily B. Fox

Stanford University

Self-Attention Long-Context Language Model Inference LLM Inference Acceleration Sparse Attention Attention Mechanisms

Apr 21, 2026cs.CV

Cognitive Alignment At No Cost: Inducing Human Attention Biases For Interpretable Vision Transformers

For state-of-the-art image understanding, Vision Transformers (ViTs) have become the standard architecture but their processing diverges substantially from human attentional characteristics. We investigate whether this cognitive gap can be shrunk by fine-tuning the self-attention weights of Google's ViT-B/16 on human saliency fixation maps. To isolate the effects of semantically relevant signals from generic human supervision, the tuned model is compared against a shuffled control. Fine-tuning significantly improved alignment across five saliency metrics and induced three hallmark human-like biases: tuning reversed the baseline's anti-human large-object bias toward small-objects, amplified the animacy preference and diminished extreme attention entropy. Bayesian parity analysis provides decisive to very-strong evidence that this cognitive alignment comes at no cost to the model's original classification performance on in- (ImageNet), corrupted (ImageNet-C) and out-of-distribution (ObjectNet) benchmarks. An equivalent procedure applied to a ResNet-50 Convolutional Neural Network (CNN) instead degraded both alignment and accuracy, suggesting that the ViT's modular self-attention mechanism is uniquely suited for dissociating spatial priority from representational logic. These findings demonstrate that biologically grounded priors can be instilled as a free emergent property of human-aligned attention, to improve transformer interpretability.

Ethan Knights

Cambridge, UK

Transformer Interpretability Visual Attention AI Alignment Vision Transformer Attention Mechanisms

Apr 21, 2026cs.CL

Tracing Relational Knowledge Recall in Large Language Models

We study how large language models recall relational knowledge during text generation, with a focus on identifying latent representations suitable for relation classification via linear probes. Prior work shows how attention heads and MLPs interact to resolve subject, predicate, and object, but it remains unclear which representations support faithful linear relation classification and why some relation types are easier to capture linearly than others. We systematically evaluate different latent representations derived from attention head and MLP contributions, showing that per-head attention contributions to the residual stream are comparatively strong features for linear relation classification. Feature attribution analyses of the trained probes, as well as characteristics of the different relation types, reveal clear correlations between probe accuracy and relation specificity, entity connectedness, and how distributed the signal on which the probe relies is across attention heads. Finally, we show how token-level feature attribution of probe predictions can be used to reveal probe behavior in further detail.

Nicholas Popovič, Michael Färber

TU Dresden & ScaDS.AI Dresden/Leipzig, Germany

Linear Probing Factual Knowledge in Language Models Attention Mechanisms Language Model Probing

Apr 21, 2026cs.LG

Super Apriel: One Checkpoint, Many Speeds

We release Super Apriel, a 15B-parameter supernet in which every decoder layer provides four trained mixer choices -- Full Attention (FA), Sliding Window Attention (SWA), Kimi Delta Attention (KDA), and Gated DeltaNet (GDN). A placement selects one mixer per layer; placements can be switched between requests at serving time without reloading weights, enabling multiple speed presets from a single checkpoint. The shared checkpoint also enables speculative decoding without a separate draft model. The all-FA preset matches the Apriel 1.6 teacher on all reported benchmarks; recommended hybrid presets span

2.9\times

to

10.7\times

decode throughput at 96% to 77% quality retention, with throughput advantages that compound at longer context lengths. With four mixer types across 48 layers, the configuration space is vast. A surrogate that predicts placement quality from the per-layer mixer assignment makes the speed-quality landscape tractable and identifies the best tradeoffs at each speed level. We investigate whether the best configurations at each speed level can be identified early in training or only after convergence. Rankings stabilize quickly at 0.5B scale, but the most efficient configurations exhibit higher instability at 15B, cautioning against extrapolation from smaller models. Super Apriel is trained by stochastic distillation from a frozen Apriel 1.6 teacher, followed by supervised fine-tuning. We release the supernet weights, Fast-LLM training code, vLLM serving code, and a placement optimization toolkit.

SLAM Labs, :, Oleksiy Ostapenko +13

SLAM Labs · ServiceNow Research · University of Toronto +4

LLM Inference Acceleration Attention Mechanisms Neural Architecture Search

Apr 21, 2026cs.CL

The Structural Attention Tax: How Retrieval Format Hijacks In-Context Learning Independent of Content

Retrieval-augmented generation (RAG) systems inject external knowledge to improve LLM outputs, yet the format of injected content -- distinct from its semantic relevance -- can independently distort the model's attention distribution. We identify and formalise a phenomenon we term the structural attention tax: knowledge graph (KG) triples, due to their relational delimiters and repeated slot patterns, capture 2-3x more attention per token than semantically equivalent natural-language text (

\hat{o}

(KG)

\approx

0.70 vs.

\hat{o}

(neutral)

\approx

0.25), compressing demonstration attention by up to 42% -- regardless of whether the triples are relevant or noise. We develop a formal framework decomposing attention scores into semantic and structural components (Eq. 2), derive a compression bound (Proposition 1) connecting token-level format bias to demonstration attention loss, and show that the structural term governs how much attention is diverted while the semantic term governs whether this helps or hurts. This decoupling reveals two orthogonal axes for improving retrieval-augmented ICL: optimising retrieval quality (semantic axis) and reducing format-driven attention capture (structural axis). Empirically, across two model families (Mistral-7B, LLaMA-3-8B) and three QA benchmarks, we observe that source-task alignment dominates: task-matched BM25 retrieval achieves 58-62% on HotpotQA vs. ConceptNet's 25-27%, a >30 pp gap that dwarfs all gating strategies (

\leq

2 pp). We derive five structure-aware mitigation strategies from the framework, ranging from zero-cost prompt modifications to training-time regularisation; format flattening (S3) is validated by both accuracy and attention-level evidence from a verbalized-triple control, while structural dispersal (S1) yields mixed results that illuminate the challenges of format-level intervention.

Yuqi Zhang, Di Zhang

Xi’an Jiaotong-Liverpool University

Retrieval-Augmented Generation Self-Attention Graph Retrieval-Augmented Generation Transformer Attention In-Context Learning Attention Mechanisms+1

Apr 21, 2026cs.LG

Nexusformer: Nonlinear Attention Expansion for Stable and Inheritable Transformer Scaling

Scaling Transformers typically necessitates training larger models from scratch, as standard architectures struggle to expand without discarding learned representations. We identify the primary bottleneck in the attention mechanism's linear projections, which strictly confine feature extraction to fixed-dimensional subspaces, limiting both expressivity and incremental capacity. To address this, we introduce Nexusformer, which replaces linear

Q/K/V

projections with a Nexus-Rank layer, a three-stage nonlinear mapping driven by dual activations in progressively higher dimensional spaces. This design overcomes the linearity constraint and enables lossless structured growth: new capacity can be injected along two axes via zero-initialized blocks that preserve pretrained knowledge. Experiments on language modeling and reasoning benchmarks demonstrate that Nexusformer matches Tokenformer's perplexity using up to 41.5% less training compute during progressive scaling (240M to 440M). Furthermore, our analysis of growth dynamics reveals that zero initialization induces a stable convergence trajectory, allowing us to derive a geometric scaling law that accurately predicts performance across expansion scales.

Weijie Zhao, Mingquan Liu, Bolun Wang +4

Luxi Tech, Shenzhen, China · The School of Computer Science, China University of Geosciences, 388 Lumo Road, Wuhan, 430074, Hubei, China.

Transformer Language Model Scaling Laws Transformer Attention Language Model Scaling Attention Mechanisms

Apr 21, 2026cs.LG

LLMs Know They're Wrong and Agree Anyway: The Shared Sycophancy-Lying Circuit

When a language model agrees with a user's false belief, is it failing to detect the error, or noticing and agreeing anyway? We show the latter. Across twelve open-weight models from five labs, spanning small to frontier scale, the same small set of attention heads carries a "this statement is wrong" signal, whether the model is evaluating a claim on its own or being pressured to agree with a user. Silencing these heads flips sycophantic behavior sharply while leaving factual accuracy intact, so the circuit controls deference rather than knowledge. Edge-level path patching confirms that the same head-to-head connections drive sycophancy, factual lying, and instructed lying. Opinion-agreement, where no factual ground truth exists, reuses these head positions but writes into an orthogonal direction, ruling out a simple "truth-direction" reading of the substrate. Alignment training leaves this circuit in place: an RLHF refresh cuts sycophantic behavior roughly tenfold while the shared heads persist or grow, a pattern that replicates on an independent model family and under targeted anti-sycophancy DPO. When these models sycophant, they register that the user is wrong and agree anyway.

Manav Pandey

Georgia Institute of Technology

LLM Sycophancy LLM Alignment Deception in Language Models Mechanistic Interpretability Attention Mechanisms Circuits in Language Models+1

Apr 21, 2026cs.LG

FG $^2$ -GDN: Enhancing Long-Context Gated Delta Networks with Doubly Fine-Grained Control

Linear attention mechanisms have emerged as promising alternatives to softmax attention, offering linear-time complexity during inference. Recent advances such as Gated DeltaNet (GDN) and Kimi Delta Attention (KDA) have demonstrated that the delta rule, an online gradient descent update, enables superior associative recall compared to simple additive updates. While KDA refined the coarse head-wise decay gate into channel-wise decay, the learning rate

β_t

in the delta update remains a scalar, limiting the model's capacity for dimension-specific adaptation. We introduce FG

^2

-GDN, which replaces the scalar

β_t

with a channel-wise vector analogous to the transition from SGD to per-coordinate adaptive optimizers such as AdaGrad and Adam. We further propose FG

^2

-GDN+, which decouples the scaling for keys and values, enabling independent control of erasure strength and write strength. Experiments on synthetic and real-world benchmarks show that FG

^2

-GDN and its variant improve associative recall and long-context understanding over GDN and KDA, with comparable computational efficiency.

Pingwei Sun, Yuxuan Hu, Jianchao Tan +6

Meituan · Beijing, China

Long-Context Modeling Linear Attention Delta-Rule Linear Attention Attention Mechanisms Gated Delta Networks Associative Memory

Apr 20, 2026cs.CV

DeltaSeg: Tiered Attention and Deep Delta Learning for Multi-Class Structural Defect Segmentation

Automated segmentation of structural defects from visual inspection imagery remains challenging due to the diversity of damage types, extreme class imbalance, and the need for precise boundary delineation. This paper presents DeltaSeg, a U-shaped encoder-decoder architecture with a tiered attention strategy that integrates Squeeze-and-Excitation (SE) channel attention in the encoder, Coordinate Attention at the bottleneck and decoder, and a novel Deep Delta Attention (DDA) mechanism in the skip connections. The encoder uses depthwise separable convolutions with dilated stages to maintain spatial resolution while expanding the receptive field. Atrous Spatial Pyramid Pooling (ASPP) at the bottleneck captures multi-scale context. The DDA module refines skip connections through a dual-path scheme combining a learned delta operator for nuisance feature suppression with spatial attention gates conditioned on decoder signals. Deep supervision through multi-scale auxiliary heads further strengthens gradient flow and encourages semantically meaningful features at intermediate decoder stages. We evaluate DeltaSeg on two datasets: the S2DS dataset (7 classes) and the Culvert-Sewer Defect Dataset (CSDD, 9 classes). Across both benchmarks, DeltaSeg consistently outperforms 12 competing architectures including U-Net, SA-UNet, UNet3+, SegFormer, Swin-UNet, EGE-UNet, FPN, and Mobile-UNETR, demonstrating strong generalization across damage types, imaging conditions, and structural geometries.

Enrique Hernandez Noguera, Md Meftahul Ferdaus, Elias Ioup +1

University of New Orleans, New Orleans, LA, USA · Center for Geospatial Sciences, Naval Research Laboratory, Hancock County, MS, USA

Attention Mechanisms Semantic Segmentation

Apr 20, 2026cs.LG

Sessa: Selective State Space Attention

Modern sequence modeling is dominated by two families: Transformers, whose self-attention can access arbitrary elements of the visible sequence, and structured state-space models, which propagate information through an explicit recurrent state. These mechanisms face different limitations on long contexts: when attention is diffuse, the influence of individual tokens is diluted across the effective support, while recurrent state propagation can lose long-range sensitivity unless information is actively preserved. As a result, both mechanisms face challenges in preserving and selectively retrieving information over long contexts. We propose Sessa, a decoder that places attention inside a recurrent feedback path. This creates many attention-based paths through which past tokens can influence future states, rather than relying on a single attention read or a single recurrent chain. We prove that, under explicit assumptions and matched regimes, Sessa admits power-law memory tails

O(\ell^{-β})

for

0 < β< 1

, with slower decay than in the corresponding Transformer and Mamba-style baselines. We further give an explicit construction that achieves this power-law rate. Under the same assumptions, Sessa is the only model class among those considered that realizes flexible selective retrieval, including profiles whose influence does not decay with distance. Consistent with this theoretical advantage, across matched experiments, Sessa achieves the strongest performance on long-context benchmarks while remaining competitive with Transformer and Mamba-style baselines on short-context language modeling.

Liubomyr Horbatko

Self-Attention Long-Context Language Modeling Persistent Memory for Language Models Recurrent Transformers Attention Mechanisms State Space Models

Apr 20, 2026cs.CV

Attention-ResUNet for Automated Fetal Head Segmentation

Automated fetal head segmentation in ultrasound images is critical for accurate biometric measurements in prenatal care. While existing deep learning approaches have achieved a reasonable performance, they struggle with issues like low contrast, noise, and complex anatomical boundaries which are inherent to ultrasound imaging. This paper presents Attention-ResUNet. It is a novel architecture that synergistically combines residual learning with multi-scale attention mechanisms in order to achieve enhanced fetal head segmentation. Our approach integrates attention gates at four decoder levels to focus selectively on anatomically relevant regions while suppressing the background noise, and complemented by residual connections which facilitates gradient flow and feature reuse. Extensive evaluation on the HC18 Challenge dataset where n = 200 demonstrates that Attention ResUNet achieves a superior performance with a mean Dice score of 99.30 +/- 0.14% against similar architectures. It significantly outperforms five baseline architectures including ResUNet (99.26%), Attention U-Net (98.79%), Swin U-Net (98.60%), Standard U-Net (98.58%), and U-Net++ (97.46%). Through statistical analysis we confirm highly significant improvements (p < 0.001) with effect sizes that range from 0.230 to 13.159 (Cohen's d). Using Saliency map analysis, we reveal that our architecture produces highly concentrated, anatomically consistent activation patterns, which demonstrate an enhanced interpretability which is crucial for clinical deployment. The proposed method establishes a new state of the art performance for automated fetal head segmentation whilst maintaining computational efficiency with 14.7M parameters and a 45 GFLOPs inference cost. Code repository: https://github.com/Ammar-ss

Ammar Bhilwarawala, Mainak Bandyopadhyay

School of Computer Engineering, KIIT Deemed to be University, Bhubaneswar, India

Ultrasound Image Segmentation Medical Image Segmentation Attention Mechanisms

Apr 20, 2026cs.AI

Stability Implies Redundancy: Delta Attention Selective Halting for Efficient Long-Context Prefilling

Prefilling computational costs pose a significant bottleneck for Large Language Models (LLMs) and Large Multimodal Models (LMMs) in long-context settings. While token pruning reduces sequence length, prior methods rely on heuristics that break compatibility with hardware-efficient kernels like FlashAttention. In this work, we observe that tokens evolve toward \textit{semantic fixing points}, making further processing redundant. To this end, we introduce Delta Attention Selective Halting (DASH), a training-free policy that monitors the layer-wise update dynamics of the self-attention mechanism to selectively halt stabilized tokens. Extensive evaluation confirms that DASH generalizes across language and vision benchmarks, delivering significant prefill speedups while preserving model accuracy and hardware efficiency. Code will be released at https://github.com/verach3n/DASH.git.

Yujie Chen, Tailai Chen, Yifeng Gao +4

Shanghai Jiao Tong University · University of California, San Diego · Carnegie Mellon University

Self-Attention Long-Context Language Model Inference Efficient VLM Inference LLM Inference Acceleration Attention Mechanisms

Apr 20, 2026cs.AI

DSAINet: An Efficient Dual-Scale Attentive Interaction Network for General EEG Decoding

In real-world applications of noninvasive electroencephalography (EEG), specialized decoders often show limited generalizability across diverse tasks under subject-independent settings. One central challenge is that task-relevant EEG signals often follow different temporal organization patterns across tasks, while many existing methods rely on task-tailored architectural designs that introduce task-specific temporal inductive biases. This mismatch makes it difficult to adapt temporal modeling across tasks without changing the model configuration. To address these challenges, we propose DSAINet, an efficient dual-scale attentive interaction network for general EEG decoding. Specifically, DSAINet constructs shared spatiotemporal token representations from raw EEG signals and models diverse temporal dynamics through parallel convolutional branches at fine and coarse scales. The resulting representations are then adaptively refined by intra-branch attention to emphasize salient scale-specific patterns and by inter-branch attention to integrate task-relevant features across scales, followed by adaptive token aggregation to yield a compact representation for prediction. Extensive experiments on five downstream EEG decoding tasks across ten public datasets show that DSAINet consistently outperforms 13 representative baselines under strict subject-independent evaluation. Notably, this performance is achieved using the same architecture hyperparameters across datasets. Moreover, DSAINet achieves a favorable accuracy-efficiency trade-off with only about 77K trainable parameters and provides interpretable neurophysiological insights. The code is publicly available at https://github.com/zy0929/DSAINet.

Zhiyuan Ma, Zeyuan Li, Zihao Qiu +6

School of Biomedical Engineering, Tsinghua Laboratory of Brain and Intelligence, and IDG/McGovern Institute for Brain Research, Tsinghua University, Beijing 100084, China · School of Basic Medical Sciences and Tsinghua Laboratory of Brain and Intelligence, Tsinghua University, Beijing 100084, China · Department of Psychological and Cognitive Sciences, Tsinghua University, Beijing 100084, China +1

Cross-Subject EEG Decoding EEG Decoding Attention Mechanisms Multiscale Time-Series Modeling

Apr 19, 2026cs.IR

HeadRank: Decoding-Free Passage Reranking via Preference-Aligned Attention Heads

Decoding-free reranking methods that read relevance signals directly from LLM attention weights offer significant latency advantages over autoregressive approaches, yet suffer from attention score homogenization: middle-context documents receive near-identical scores, destroying the fine-grained distinctions required for ranking. We propose HeadRank, a framework that lifts preference optimization from discrete token space into the continuous attention domain through entropy-regularized head selection, hard adjacent-level preference pairs, and a distribution regularizer that jointly sharpen discriminability in the homogenized middle zone. Depth truncation at the deepest selected layer further reduces inference to

\mathcal{O}(1)

forward passes. Across 14 benchmarks on three Qwen3 scales (0.6B--4B) using only 211 training queries, HeadRank achieves the highest average NDCG@10 at every scale, outperforming both generative and decoding-free baselines on the majority of benchmarks with 100% formatting success. At 4B, 57.4% of relevant middle-zone documents reach the top quartile versus 14.2% for irrelevant ones -- a 43-percentage-point selectivity gap that demonstrates the effectiveness of attention-space preference alignment for listwise reranking.

Juyuan Wang, Chenxing Wang, Yuchen Fang +8

Weixin Group, Tencent, China · 2South China University of Technology, Guangzhou, China

Attention Head Analysis Self-Attention Learning to Rank LLM Reranking Attention Mechanisms Preference Alignment+1