Vision Transformer

Vision Transformers (ViTs) are difficult to interpret because current methods of relevance propagation and attention flow do not fully consider some key architectural features, such as the uneven importance of attention heads and residual connections. Prior approaches typically assume uniform importance across attention heads; furthermore, they model skip connections as identity paths, leading to inaccurate relevance attribution. To address these issues, we introduce GradSkip, a novel relevance propagation method for ViTs based on adaptive head weighting and skip-aware propagation. GradSkip models the different importance of the attention heads and dynamically distributes relevance between the attention and residual paths. Experiments on ImageNet1K and BloodMNIST demonstrate a state-of-the-art faithfulness of GradSkip while requiring over 14 times fewer GFLOPs than the best-performing existing approaches. Additional evaluations using transformer-based segmentation confirm improved localization and alignment with ground-truth regions.

Figure-ground organization in the human visual system relies on several shape-based cues, including surroundedness, convexity, and symmetry. While these cues have been extensively studied using abstract stimuli, little is known about how they operate under natural conditions or how they arise from the statistics of natural scenes. Deep neural networks offer a promising path forward: a model that relies on the same figure-ground cues as humans would provide tractable experimental access to the underlying mechanisms. In this study, we evaluate shape-based figure-ground organization in Vision Transformers (ViTs), for which prior work has demonstrated the emergence of object-based grouping. We test 25 ViTs spanning supervised and self-supervised training objectives, by fitting linear probes to predict figure-ground assignment from intermediate patch representations using both natural images and controlled artificial stimuli that isolate individual cues. Our results show that ViTs robustly encode surroundedness and convexity, and that probes trained on natural images generalize zero-shot to artificial stimuli across several models. For symmetry we observe mixed results: the cue is encoded for uniformly colored but not for textured regions. Taken together, our findings demonstrate that Gestalt-like figure-ground cues can be learned from natural scene statistics and position ViTs as a compelling model system for studying the computational mechanisms of perceptual organization. Code and data is available at https://github.com/mtangemann/mlvbench

In computational vision science, Convolutional Neural Networks (CNNs) have emerged as a popular model of biological vision because of the alignment they can exhibit with neural and behavioral data in humans and animals. However, it remains unclear to what extent this alignment persists for visual tasks that extend beyond the canonical object recognition paradigm based on well defined semantic content. In this study, we diverge from the common object-centric view by focusing on another aspect of vision: texture perception. We consider textures of different complexity generated with three different algorithms from the same source images. Using a rank-based statistic, we quantify the information encoded in the internal representations of a CNN and three Vision Transformers (ViTs), and we compare the similarity of these representations to those inferred from human psychophysics data. We find that the representation of textures is aligned in different ViTs, but not between the ViTs and the CNN; that ViTs form similar representations for textures of different complexity; that human performance in recognizing textures can be better predicted from ViTs representations rather than CNN representations. Taken together, these results suggest that ViTs may capture more faithfully than CNNs how texture patterns are visually processed by humans, and that the representations of texture stimuli in computational models may be driven by the network architecture.

Vision Transformers (ViTs) remain vulnerable to localized adversarial attacks, e.g., adversarial patches, while recent test-time defenses mitigate them by suppressing image tokens with abnormally high attention scores. These defenses exploit a strong coupling between attention and adversarial effectiveness: adversarial tokens often need to attract substantial attention to influence the prediction. We introduce adversarial decoys, independently optimized image patches that redirect the attention, and therefore related defenses, toward selected target tokens. Rather than jointly optimizing misclassifications and defense evasion, our approach decouples the two objectives: the original adversarial region induces the incorrect prediction, while a separate decoy manipulates the attention ranking used by the defense. A layer-wise objective increases target-token attention and promotes these tokens above competing non-target ones. Since the decoy is optimized independently of the underlying attack, the method is attack-agnostic and can be easily integrated with any existing adversarial patch attack. Experiments on ImageNet across multiple ViT architectures and attacks show that decoys can redirect high attention scores away from the true adversarial region while preserving much of the attack effectiveness. These results reveal a fundamental limitation of using attention magnitude as an indicator of adversarial relevance.

Accurate tumour localization and diagnosis is a critical component of clinical care for brain cancers. Magnetic Resonance Imaging (MRI) is the most commonly used imaging modality due to its superior soft-tissue contrast. However, standard MRI often exhibits limited contrast and imaging artifacts, which necessitates the use of contrast agents to enhance lesion visibility. The administration of chemical contrast agents is not always feasible and may be contraindicated in patients with renal impairment or other health conditions. As a result, developing accurate and non-invasive contrast enhanced MRI (CEMRI) synthesis methods has clinical importance. In recent years, numerous approaches for CEMRI synthesis have been proposed, predominantly relying on generative artificial intelligence models. While these methods demonstrate promising performance, their dependence on implicit feature learning often limits their ability to preserve anatomical boundaries and tumour-specific fine structures. To address these challenges, we propose an anatomically aware frequency-and-structure-guided vision transformer (AA-ViT), for CEMRI synthesis using pre-contrast MRI modalities (T1, T2, and FLAIR). Experiments on the BraTS 2021 dataset demonstrate that the proposed method preserves anatomical and lesion boundaries, achieving higher PSNR and SSIM than state-of-the-art approaches. Clinical evaluation by three neuroradiologists and a neurosurgeon on 19 randomly selected cases across diverse gliomas yielded a mean score of 3.94/5, providing preliminary clinical validation rarely seen in prior studies. Synthetic post-contrast scans from our model could lower scanning costs, shorten imaging time, and avoid the potential risks of using gadolinium-based contrast agents.

Memory-efficient optimizers such as GaLore train large language models by projecting gradients onto a rank-r subspace recomputed every T steps, assuming this subspace is a slowly drifting object that can be tracked. We show that beyond a small reproducible core, there is no such object. Two estimates of the top-r subspace computed at the same step from disjoint minibatches disagree as much as estimates computed T steps apart (0.73 vs 0.74 of the maximal chordal distance sqrt(2r), at Pythia-160M with r=128): the apparent rotation at each refresh is dominated by estimator noise. This holds across four model families in three architecture classes from 70M to 6.9B parameters, strengthening with scale, and more weakly in a vision transformer. Only ~39 of 128 directions are reproducible across minibatches, and averaging cannot recover the rest: under N-fold averaging the gradient's spectral tail shrinks as N^(-1/4) rather than the N^(-1/2) of pure noise, so no averaging budget makes the subspace well defined. What helps instead follows from treating each refresh as a change of coordinates for Adam's state. Carrying the second moment blindly is provably about (r-k*)/2 worse than the best rotation-blind estimator, while the first moment transports exactly through the rotation, the optimal linear map under isotropic gradients and the rule LDAdam uses. At 1B over 40k steps (3 seeds), full LDAdam reaches 18.7 perplexity at beta2=0.999, beating untransported GaLore after its best beta2 fix (19.3); shortening the second-moment memory to beta2=0.99 helps the refreshing optimizers, though for canonical GaLore the effect is small and a full-rank control reverses it. One measurable fact, subspace non-identifiability, clarifies why GaLore works, which patches work, and what to check before trusting a low-rank assumption: the reproducible rank k*.

Convolutional Neural Network (CNN) and Vision Transformer (ViT) for image classification exploit a dense grid of pixels containing redundant information. Consequently, for a larger image dataset, CNNs and ViTs face deployability challenges due to high computational complexity. Representing images as graphs of superpixels offers an efficient alternative that preserves key information while eliminating pixel-level redundancy. Graph Neural Networks (GNNs) have been utilized on such graphs to perform image classification. However, GNNs are known to struggle with capturing long-range dependencies which is important in the domain of image classification. Furthermore, a majority of these superpixel-based image classification approaches do not explicitly preserve translation/rotation invariance. Nevertheless, preserving translation/rotation invariance is important for robust image classification. Thus, this paper proposes SuperGT, a Graph Transformer-based framework for image classification, which captures the long range dependencies, along with a pre-processing scheme that preserves translation/rotation invariance. We evaluate SuperGT on CIFAR-10 dataset and observe that it performs significantly better than many baselines. Furthermore, we note that the overall performance of SuperGT is comparable to the previous state-of-the-art model, namely, ShapeGNN, without relying on coordinates of the boundary points of each superpixel required by ShapeGNN.

While prior studies have successfully compressed vision Transformers (ViTs) through various pruning techniques, most have concentrated on width pruning to achieve significant reductions in model size. Depth pruning, which removes entire layers from a ViT, is notoriously difficult for accuracy recovery despite its potential to deliver higher speedups, limiting the acceleration achieved by existing joint width-and-depth pruning methods. In this work, we reveal that the failure of existing depth pruning methods lies in their neglect of heterogeneity between different layers, and we introduce HetDPT, a heterogeneity-aware depth pruning method that avoids dimension mismatch. Comprehensive experiments on ImageNet-1K, CIFAR-100, COCO, and ADE20K validate our method: HetDPT achieves a 1.58$\times$ speedup for DeiT-B while maintaining accuracy and a 1.39$\times$ speedup for DeiT-S with nearly no accuracy degradation. Furthermore, when combined with width pruning, HetDPT+ sets a new state-of-the-art record in extreme ViT pruning, enhancing the acceleration ratio from 4.24$\times$ to 5.19$\times$ for the Isomorphic-Pruning-2.6G configuration while maintaining near-lossless accuracy; our code is available at https://github.com/Efficient-AI-for-All/HetDPT.

Traditional malware detection methods struggle to generalize to obfuscated or previously unseen threats. This paper introduces ThreatVisionAI, a hybrid malware family classification framework that integrates a raw-image CNN, a wavelet-based CNN, and a Vision Transformer (ViT) to capture complementary spatial, frequency-domain, and global relational features in malware images. The wavelet-based CNN captures multi-scale frequency information that helps distinguish closely related families, while the ViT branch models long-range dependencies across the image. Evaluated on the Malimg dataset, ThreatVisionAI achieves 98.01% accuracy and a weighted F1 score of 0.9742, with wavelet-domain features providing measurable gains on minority and visually similar families. These results confirm that frequency-aware and transformer-based representations improve image-based malware family classification.

Diffusion architectures now encompass convolutional UNets as well as transformer-based designs such as Diffusion Transformers (DiTs), inspired by Vision Transformers (ViTs), yet the effects of structured geometric perturbations within these architectures remain poorly understood. We study this question through a unified framework that applies reflection-based elements of the dihedral group to intermediate hidden states as controlled internal interventions, contrasting geometrically consistent and inconsistent variants. Using activation-level diagnostics, including Self-Consistency Shift (SCS), Activation Mass Scatter (AMS), and Drift, we analyze feature stability and geometric drift. We find that consistent transformations improve stability, while inconsistent ones induce predictable, architecture-specific failures. In the main Stable Diffusion 2.1 U-Net study, we evaluate seven intervention modes over three seeds and complement the internal diagnostics with image-level FID, KID, CLIP score, and LPIPS diversity. Taken together with supporting ViT and controlled DiT analyses, these results establish geometric consistency as a key principle for stable hidden-state interventions in spatially structured vision and diffusion models.

Glaucoma is a progressive eye disease that can lead to irreversible vision loss if not detected at an early stage. Conventional diagnostic procedures are often time-consuming and rely heavily on expert interpretation, limiting their scalability for large-scale screening. In this study, glaucoma detection is investigated under two evaluation settings: sample-wise, where individual samples are analyzed independently, and patient-wise, where data from each patient are aggregated for final prediction. An automated multimodal framework is proposed that integrates fundus images with clinical data. Under the sample-wise setting, detection is performed using fundus images and clinical features individually, as well as through their multimodal combination. Under the patient-wise setting, predictions are obtained by aggregating multiple fundus image representations with corresponding clinical information for each patient. Deep visual features are extracted using a Vision Transformer (ViT) architecture and classified using classical machine-learning models, with a stacking-based ensemble of the three best-performing classifiers employed to optimize performance. Experiments conducted on the publicly available PAPILA dataset demonstrate strong diagnostic performance, achieving 97.47% accuracy and a 97.50% F1-score for sample-wise multimodal classification, and 98.97% accuracy and F1-score for subject-wise detection. The proposed framework is further deployed as an end-to-end web-based platform to support automated glaucoma screening and clinical decision support.

Models trained via Contrastive Language-Image Pretraining (CLIP) serve as the foundational vision encoders for most modern Large Vision Language Models (LVLMs). Despite their widespread adoption, CLIP models exhibit a critical yet underexplored failure mode: irrelevant text appearing within images confounds visual representations, biasing them toward lexical meaning rather than true visual semantics. This robustness issue, commonly described as a Typographic Attack (TA), exposes a vulnerability that poses a significant risk to safety-critical applications such as autonomous driving. To achieve interpretable and effective robustness against TA, we propose a novel, training-free mechanistic interpretability method. Our method provides sampling-based interpretations of hidden state representations and quantitatively attributes semantic versus lexical focus to individual attention heads. Through probabilistic analysis and circuit mining, we isolate specific Vision Transformer (ViT) components that disproportionately encode lexical information, thereby identifying the mechanistic source of TA. We further show that simple interventions applied directly to the identified circuits, without any additional training, can substantially improve robustness against Typographic Attacks in object classification. These interventions, such as selective adjustment of attention weights, also outperform both supervised and training-free defense methods. Our experiments demonstrate that applying the proposed intervention to the vision encoders of several state-of-the-art LVLMs yields substantial gains in Visual Question Answering accuracy under Typographic Attack interference on RIO-Bench. These results confirm both the efficacy and the generalizability of our mechanistic approach. Code is released at https://github.com/Liu-524/SamplingTAR.

While Vision Transformers have achieved remarkable success across computer vision and language applications, the geometric evolution of their internal representations throughout training remains insufficiently understood. Existing analyses primarily focus on attention mechanisms and downstream performance, leaving the evolution of representation geometry largely unexplored. In this work, we present Transformer Geometry Observatory-II (TGO-II), a representation geometry analysis framework designed to investigate how Transformer representations evolve during supervised training. TGO-II analyzes Vision Transformer (ViT-Small/16) representations using Centered Kernel Alignment (CKA), Singular Vector Canonical Correlation Analysis (SVCCA), Two-Nearest Neighbor Intrinsic Dimensionality (TwoNN-ID), and token covariance analysis. Our experiments reveal three key observations. First, both CKA and SVCCA progressively decrease throughout training, indicating increasing representational specialization across Transformer layers. Second, intrinsic dimensionality consistently increases before stabilizing, suggesting progressive expansion of the representation manifold into a larger set of locally accessible degrees of freedom. Third, token covariance and coupling analyses demonstrate that strong token interaction structure persists throughout training, challenging the hypothesis that increasing representational complexity arises primarily from progressive token independence. These findings suggest that representation complexity and layer specialization emerge simultaneously during training. Manifold expansion appears to occur without token decoupling. Together, these observations motivate a new hypothesis in which Vision Transformers increase representational complexity through progressively richer transformations while preserving strong token interaction structure during learning.

We introduce a controlled subspace intervention framework to investigate how self-supervised Vision Transformers (ViTs) encode dense geometric information. While linear probing is widely used to assess geometric representations, it treats features as a black box, failing to disentangle the underlying topology. To address this issue, we decompose the weights of converged linear probes to isolate the low-rank subspaces containing explicit geometric signals using Singular Value Decomposition (SVD). Our perspective yields three key insights: (1) Pre-training objectives determine how features are encoded. DINOv2 aligns spatial features for efficient linear extraction, while Masked Autoencoders (MAE) tend to disperse these signals, requiring a broader spatial context. (2) Explicit geometric representations are highly compressible, suggesting dense predictive heads could potentially be constrained to low-rank subspaces with minimal performance loss. (3) The layer-wise task affinity suggests that geometric precision peaks at intermediate layers before yielding to semantic abstraction in the final layers. By connecting internal encoding mechanics with downstream performance, these findings provide a basis for effective feature selection and lightweight decoder design. The source code is available at https://github.com/Zhou-Weichen/Geosubprobe.

Dual-pixel (DP) imaging enables metric depth estimation from a single camera using sub-aperture disparity. However, the extremely small effective baseline limits disparity observability, leading to structural degradation and depth failure in textureless, low-contrast, or downsampled regions. Existing DP-based methods rely primarily on local disparity cues and therefore become unreliable when disparity signals are weak or ambiguous. To address this limitation, we propose \emph{FoundDP}, a unified framework that integrates metric DP depth with global structural priors from a monocular depth foundation model. Our method preserves metric scale through DP-derived depth and leverages Vision Transformer (ViT) features to restore structural consistency in weak-disparity regions. To ensure reliable metric guidance under DP imaging conditions, we identify and mitigate ViT representation degradation induced by DP defocus blur via ViT feature alignment, enabling stable metric-guided depth estimation. Extensive experiments on synthetic and real-world DP benchmarks show that FoundDP delivers superior performance, with consistent gains in structural fidelity and metric accuracy, especially under reduced disparity observability. Code will be available at: https://github.com/EchoLighting/FoundDP

Vision Transformers achieve strong image classification accuracy but process all image regions with nearly the same computation, even when many regions are redundant or uninformative. Recent adaptive inference methods reduce this cost by selectively compressing tokens or terminating inference early, but combining these mechanisms often causes unstable intermediate representations and accuracy degradation. We introduce Fusion, a unified adaptive inference framework that coordinates token merging, early exiting, and token pruning through a simple staged design: tokens are merged first, confidence is evaluated next, and pruning is applied only to samples that continue inference. This ordering allows the three mechanisms to operate cooperatively rather than competitively. Fusion further includes lightweight routing modules that adapt compression strength to each input and support inference-time adjustment of the accuracy--latency trade-off without retraining. On ImageNet-1k with DeiT-S, Fusion matches or surpasses state-of-the-art adaptive ViT methods at comparable compute budgets while reducing calibration error by up to $4\times$ and inference energy by $48\%$. Experiments across ImageNet-100, CIFAR-100, and ImageNette with multiple ViT backbones demonstrate consistent transferability without dataset-specific tuning.

Recent recursive Transformer studies have primarily reused shared parameters across computation steps to construct compact, parameter-efficient models. In this work, we leverage recursion to build effectively deeper Transformers with stronger representational capacity. However, in Vision Transformers, simply increasing recursion depth does not reliably improve performance, as existing recursive approaches do not fully utilize the intermediate representations produced throughout recursive computation. We propose Soft Mixture-of-Recursions (SoftMoR) and its Vision Transformer instantiation, Soft Recursive Vision Transformer (SR-ViT). SoftMoR learns token-wise mixture weights to softly combine outputs from all recursion steps, allowing intermediate representations to be utilized in a learnable and flexible way. Across diverse vision tasks, SR-ViT consistently improves as recursion depth increases with minimal parameter overhead. On ImageNet-1K, increasing recursion depth from 1 to 4 improves SR-ViT-S top-1 accuracy from 79.83% to 82.48% with only 1.7M additional parameters, outperforming the substantially larger DeiT-B while using approximately 27% of its parameters. These results demonstrate that SoftMoR provides a parameter-efficient path to deeper and stronger Vision Transformers through recursion.

Vision Transformers (ViTs) commonly rely on injected positional mechanisms to address self-attention's permutation invariance. Motivated by the spatial regularities of natural images, we ask whether spatial organization can be induced from data rather than explicitly injected. Under controlled, matched from-scratch training, we propose Active Spatial Guidance (Guidance), a training-only objective that disables positional injection and applies an auxiliary 2D coordinate-regression loss to the final-layer patch tokens. The guidance head is used only during training and removed for inference; the deployed model consists of a positional-injection-free ViT encoder and the task-specific prediction module. Using DINOv3 ViT backbones, Guidance consistently improves performance on ImageNet-100 classification, ADE20K semantic segmentation, and Hypersim monocular depth estimation, outperforming strong injected baselines such as learned absolute positional embeddings and rotary positional embeddings under identical training protocols. On ImageNet-100, broader comparisons against representative injected positional designs further support Guidance's effectiveness. Guidance also improves robustness under resolution transfer, and multi-resolution training further strengthens accuracy across input sizes. Overall, our results suggest that spatial inductive bias in ViTs need not be architecturally injected, but can be shaped through training-time supervision. The code used for training and evaluation is publicly available in https://github.com/cloudlc/asg.

Medical image segmentation is dominated by U-Net-style encoder-decoder architectures. Vision Transformers (ViTs) overcome the limited receptive field of convolutional networks through self-attention, enabling modeling of long-range dependencies. Early ViT-based segmentation methods typically retained U-Net-style decoders because pretrained ViT representations were insufficient to support accurate dense prediction. Recent advances in large-scale pretraining have redefined the representation capability of ViTs, reducing the reliance on U-Net-style decoder architectures in modern vision models. This prompts two questions: Is the U-Net paradigm still necessary for medical image segmentation? If not, how should an encoder-only segmentation framework be designed? Motivated by these questions, we explore key architectural choices for encoder-only medical image segmentation based on modern ViT backbones and establish a query-based encoder-only design with multi-level query modeling and learnable block fusion, realized in Encoder-only Segmentation (EoSeg). Extensive experiments across seven benchmark datasets spanning CT, MRI, histopathology, endoscopy, and dermoscopy validate the effectiveness of the proposed design across diverse medical imaging modalities, including mDice scores of 85.50% on Synapse, 91.73% on ACDC, and 93.27% on GlaS. The results demonstrate that a U-Net-style decoder is no longer necessary for medical image segmentation with modern ViT backbones and further show that EoSeg provides an effective encoder-only design. Code is available at: https://github.com/Retinal-Research/EoSeg

Most token reduction methods for Vision Transformers seek favorable tradeoffs between accuracy and efficiency by pruning, merging, or pooling patch tokens. REDI (Relevance for DINOv3 Token Reduction) studies this question through a controlled supervised reference: how should a fixed token budget be allocated across patches for image classification? REDI quantizes final block DINOv3 patch representations into a visual vocabulary and derives class conditioned corpus scores using supervised TF-IDF over visual words. For each validation image, the ground truth class selects a row of the TF-IDF table, and four transformed views produce a TF-IDF map aligned to a reference center crop. A separate dense pass on the same crop provides an attention map. After independent min max normalization, their elementwise product defines the REDI score. A fixed keep, merge, and compress operator then uses score rank to assign patch roles and score magnitude to weight merging and compression. With precomputed REDI scores, a frozen DINOv3 ViT-B/16 backbone, and the same linear classifier used for dense evaluation, the operator reduces the sequence length from 201 to 107 tokens, a 46.8% sequence reduction. The REDI variant based on incoming attention mass achieves 84.706% Top-1 accuracy on ImageNet-1K, compared with 83.514% for the dense baseline, 82.634% for incoming attention mass alone, and 81.796% for supervised TF-IDF alone. The same corpus term also improves reduced classification for three alternative attention formulations relative to their attention only counterparts. Together, these controlled comparisons indicate that class specific corpus statistics and image specific attention provide complementary signals for patch ranking in this setting.

Self-supervised learning enables audio representations that transfer across domains and tasks. We present BEST-RQ-2, an evolution of BEST-RQ that retains frozen randomprojection-based discrete targets while introducing a two-step contextualize-then-predict pretraining scheme. A ViT context encoder processes only the unmasked spectrogram regions, and a lightweight predictor infers targets for the masked regions; the predictor is discarded after pretraining. Replacing the original Conformer encoder with a ViT shifts performance across domains, slightly reducing speech performance while improving music and environmental sounds, with comparable average scores. The main improvement comes from decomposing masked prediction into separate contextualization and prediction stages. On the X-ARES and XARES-LLM benchmarks, BEST-RQ-2 consistently outperforms one-stage baselines in overall transfer while keeping inference compute unchanged. Code and model checkpoints are publicly available.

Self-supervised learning (SSL) currently drives state-of-the-art performance in no-reference image quality assessment (NR-IQA). However, standard SSL pipelines uniformly apply synthetic distortions across the entire image field, which can limit their sensitivity to spatially localized and co-occurring degradations encountered in real-world content. In this work, we empirically expose this representational blind spot across existing state-of-the-art encoders, demonstrating their reduced sensitivity to spatially bounded image degradations. To bridge this gap, we introduce Spatial Localized Image Degradation Embeddings for Image Quality Assessment (SLIDE-IQA). SLIDE-IQA employs a dual-branch Vision Transformer framework that injects spatially bounded degradations into a contrastive pretraining objective. To handle the spatial complexity of these degradations, we introduce a Threshold-Bounded Exclusion Mechanism, a representational design choice that resolves structural conflicts arising from spatially localized distortions to ensure the latent space respects both degradation type and spatial scale. Finally, we show that SLIDE-IQA's synthetic-only pretraining significantly improves sensitivity to localized distortions, while achieving competitive performance on NR-IQA benchmarks against existing SSL NR-IQA models.

We present CLEAR-MoE, a four-phase post-training pipeline that converts a frozen pretrained Vision Transformer (ViT) into a sparse Mixture-of-Experts (MoE) model without updating backbone weights. The pipeline (i) scores feed-forward network (FFN) layers by sparsity, clusterability, and output sensitivity; (ii) decomposes selected layers into a shared low-rank SVD basis and per-cluster residual experts using k-means clustering; (iii) trains lightweight routers supervised by cluster labels; and (iv) dispatches tokens through pluggable CUDA backends. On Imagenette with DeiT-Small, CLEAR-MoE retains 99.9% of the dense model's accuracy (86.70 +/- 0.02% versus 86.73%). Extensive ablation studies reveal a consistent empirical finding: the shared SVD basis is the primary factor responsible for preserving accuracy. Random routing, learned routing, and three different router architectures produce nearly identical performance, with accuracy varying by at most 0.06 percentage points (86.62%-86.68%). Accuracy also remains stable across different SVD ranks, expert counts (2-8), calibration set sizes (50-500), and random seeds. This behavior generalizes across five ViT backbones (DeiT-Tiny, DeiT-Small, DeiT-Base, ViT-Small, and ViT-Base), covering models from 5.7M to 86.6M parameters, with accuracy differences <= 0.10 percentage points from their dense counterparts. On a GTX 960 GPU, routing and scatter-gather overhead make the CLEAR-MoE FFN 1.3-1.7x slower than the dense implementation. A dispatch microbenchmark further shows that routing is an order of magnitude more memory-bound than expert matrix multiplications, identifying fused dispatch kernels as a promising direction for future optimization.

Vision transformers have become a dominant architecture for visual recognition. However, standard models do not explicitly encode the planar symmetries that arise in many vision domains. We introduce a family of vision transformers equivariant to arbitrary discrete subgroups of $\mathrm{O}(2)$, providing a unified framework that generalizes prior flipping- and $D_4$-equivariant transformer architectures. Our construction yields equivariant analogues of the core transformer components, together with expressivity guarantees for the resulting layers. In particular, we show that whenever $H \le G$, the class of $G$-equivariant ViTs embeds naturally into the class of $H$-equivariant ViTs. We also prove that, in the single-head setting, the corresponding equivariant self-attention layer realizes every $G$-equivariant self-attention map representable by ordinary self-attention. We further construct a $D_6$-equivariant model based on hexagonal patches, making the architecture compatible with six-fold rotational symmetries. We evaluate the resulting models on the PatternNet aerial image dataset in artificially data-scarce regimes across subgroups of $D_4$ and $D_6$. Our experiments compare two equivariant attention mechanisms and analyze how the choice of homogeneous-space configurations used in the nonlinearities affects performance. Preliminary results under matched parameter budgets indicate that equivariance can improve recognition accuracy, motivating further study of how discrete symmetry groups shape transformer-based visual recognition models.

Recent advancements in synthetic data technology have opened a new era where images of remarkable quality are generated, blurring the lines between real-life images and those produced by Artificial Intelligence (AI). This evolution poses a significant challenge to ensuring the reliability and authenticity of data, underscoring the need for robust detection methods. In this paper, we present a robust approach aimed at addressing these pressing concerns. Our methodology revolves around leveraging fusion strategies, combining the strengths of multiple detection methods for identifying AI-generated images. Through extensive experimentation on the CIFAKE dataset, our model showcases remarkable performance, achieving an impressive accuracy rate of 97.32%. This accomplishment underscores the efficacy of our approach in accurately distinguishing between AI-generated images and real-life images, thus contributing to the advancement of data authentication techniques amidst the proliferation of synthetic data.

Few-shot image recognition requires models to adapt to new classes from a small labeled support set. Hebbian fast-weight memory can provide temporary associative information during an episode, but fixed memory behavior may not be appropriate for every few-shot task. In this work, we propose Adaptive Hebbian Routing for few-shot Vision Transformers. The method uses a lightweight MLP router to control the contribution of Hebbian memory, the strength of memory updates, and the retention of previous memory from support-set features. We study Adaptive Placement, Adaptive Plasticity, and Fully Adaptive Hebbian Routing. Experiments use ViT-Small, DeiT-Small, and Swin-Tiny under 5-way 1-shot evaluation on Omniglot, CIFAR-FS, and cross-domain transfer from CIFAR-FS to Omniglot. In the direct Swin comparison, fixed and adaptive Hebbian variants use the same memory location. Adaptive Plasticity improves the fixed Hebbian result from 96.74% to 96.92%, while Fully Adaptive Routing achieves the best result at 96.94%. The fully adaptive Swin model also reduces inference time from 16.51 ms to 14.05 ms relative to fixed Hebbian Swin. On CIFAR-FS, adaptive variants improve performance across all three backbones, and the multi-shot evaluation shows that these gains remain useful as the number of support examples increases. These results show that adaptive plasticity and adaptive memory activation can improve few-shot Transformer representations beyond fixed Hebbian behavior.

Modern VLMs and VLA systems commonly adopt off-the-shelf ViTs such as SigLIP2 as visual encoders, but diverse downstream requirements in latency, temporal modeling, and VLM integration often call for customized SOTA-level ViTs. Training such encoders remains beyond the reach of much of the community, as it requires massive image-text data, while standard softmax attention makes high-resolution or dynamic-resolution pretraining prohibitively costly and often forces low-resolution pretraining followed by post-hoc adaptation. TuringViT addresses these challenges with three key designs: Turing Linear Attention (TLA) for efficient sequence modeling, VISTA-Curation to construct supervision-rich image-video training data, and native dynamic-resolution pretraining that supports flexible inputs from the start and transfers seamlessly to downstream VLMs. As a result, TuringViT outperforms leading open-source ViT baselines with only 10% of the data, achieves stronger downstream VLM performance, and delivers substantially better latency scaling on high-resolution inputs. Our scaling-law analysis further shows that TuringViT continues to improve predictably with curated data scale, far from saturation. Its fast adaptation, hardware-friendly design, and efficient deployment have made it a unified visual foundation across XPeng's AI systems. More broadly, TuringViT provides a reproducible pipeline that dramatically lowers the cost for the community to train, customize, and deploy SOTA-level ViTs, moving toward making such Vision Transformers accessible to all.

Vision Transformers (ViT) dominate computer vision. However, their reliance on rigid patch projectors hinders transfer to Earth Observation (EO), where input modalities, scales, and resolutions vary widely. We introduce UniverSat, a ViT-style backbone built around a Universal Patch Encoder that maps patches from arbitrary spatial, spectral, and temporal resolutions, and from both optical and non-optical sensors, into a shared embedding space with a shared set of weights. This enables training a single model on heterogeneous multimodal corpora via self-supervision, yielding robust, sensor-agnostic spatial features. We validate this approach with strong results across classification and segmentation on standard EO benchmarks from GeoBench, PANGEABench, and SpectralEarth. Our code and models are available at https://github.com/gastruc/UniverSat.

The clinical diagnosis of skin diseases is susceptible to interference from inter-class similarity of skin lesions, and over-reliance on clinicians'experience easily leads to subjective bias. Although existing deep learning aided diagnosis methods achieve competitive accuracy, they suffer from the black-box opacity of Vision Transformer (ViT) and poor adaptability to medical few-shot scenarios. Moreover, mainstream explainable algorithms generally face the bottleneck of significant accuracy degradation when improving interpretability. This paper proposes an interpretable ViT (IViT) constrained by Quadratic Programming (QP). The introduced pre-trained transfer learning adapts to few-shot feature extraction. A discrete QP feature selection framework is constructed to screen generic and discriminative features consistent with clinical diagnostic logic. A multi-objective loss function is designed to reduce feature redundancy and optimize activation distribution while preserving classification performance. Experimental results on six standard skin disease datasets show that IViT achieves an accuracy of 93.80%, only 0.21% lower than the baseline, with feature redundancy reduced by 29.5%. Its core activation regions are consistent with clinically concerned lesion areas. The proposed model balances accuracy and interpretability, providing a reliable solution for the clinical deployment of few-shot intelligent skin disease diagnosis.

Parameter-efficient fine-tuning (PEFT) has become a practical solution for adapting large pretrained vision transformers (ViTs) to downstream tasks while updating only a small subset of parameters. However, existing adapter-based methods perform adaptation independently for each token, implicitly assuming that token refinements should be learned in isolation. This token-wise formulation overlooks the structured relationships among tokens that naturally arise in visual scenes, potentially leading to redundant updates and spatially inconsistent feature refinement. In this work, we revisit the design of parameter-efficient adapters and propose to perform adaptation in hyperedge space rather than token space. We introduce HyperAdapter, a hypergraph-based adapter architecture that enables structured, group-aware adaptation through soft token routing. HyperAdapter constructs a soft hypergraph over ViT tokens using prototype-based assignments, aggregates token features into latent hyperedge representations, applies lightweight bottleneck adaptation at the hyperedge level, and diffuses the resulting updates back to tokens via the hypergraph incidence structure. This design injects an explicit structural inductive bias into PEFT while preserving the modularity and efficiency of standard adapters. Extensive experiments across diverse visual benchmarks demonstrate that structured hyperedge adaptation consistently outperforms strong PEFT baselines under comparable parameter budgets, with particularly pronounced gains on tasks requiring structured reasoning. Our results suggest that the choice of adaptation space is a critical yet underexplored dimension in parameter-efficient transfer for ViTs.

Vision Transformers have achieved remarkable success in many fields, yet their deployment on edge devices remains challenging due to their substantial computational demands. Post-Training Quantization (PTQ) offers an attractive solution by compressing models using a small calibration set with minimal training overhead. However, most existing PTQ works adopt a static quantization paradigm that is uniformly applied to all instances. Given the substantial diversity of natural images, the activation distributions vary significantly across samples, making these methods inherently suboptimal. In this paper, we propose ScalePredictor, a dynamic quantization framework for accurate and efficient quantization scale learning of ViTs. We first reveal a hidden correlation between the distribution range of shallow-layer activations and the optimal scales of deeper layers. Based on this, we develop a scale learning mechanism that integrates an efficient range extraction approach to capture robust range statistics at the shallow stage, which are then fed into a Taylor-motivated polynomial scale projection module to generate all quantization scales simultaneously. With the efficiency of polynomial approximation, ScalePredictor introduces insignificant computational overhead while avoiding costly just-in-time calibration. Extensive experiments on ImageNet demonstrate that ScalePredictor consistently outperforms prior PTQ methods, achieving a more favorable accuracy-efficiency trade-off. Code and additional results are shown in the supplementary materials.

Student-generated drawings are widely used in science education to assess learners' conceptual understanding in modeling-based tasks aligned with the Next Generation Science Standards (NGSS). However, scoring such drawings requires expert human judgment to interpret complex visual representations, making large-scale assessment costly to implement and sustain in classroom settings. In this work, we study automated scoring of student-generated scientific drawings using a vision-based model. We evaluate a Vision Transformer (ViT) with parameter-efficient adaptation and propose a confidence-aware scoring framework that derives response-level confidence from test-time predictive distributions. This confidence signal enables selective automation by scoring high-confidence responses automatically while deferring uncertain cases for human review. Experiments on six NGSS-aligned middle school assessment items show that the proposed approach improves scoring reliability while supporting a practical trade-off between automated coverage and scoring risk, highlighting the value of confidence-aware methods for trustworthy educational assessment.

Deploying vision transformers (ViTs) on sensor-edge systems is limited not only by on-device compute, but also by the energy and bandwidth required to transmit high-dimensional image data from the sensor to the processor. While in-sensor and near-sensor computing reduce this cost through early feature extraction, existing methods often provide only modest compression. We observe that the frequency domain provides a naturally compact representation of visual information and can be exploited at the sensor level to reduce sensor-to-processor data movement. Building on this insight, we present FrequencyFormer, a co-designed sensor-to-processor pipeline for efficient ViT inference. FrequencyFormer includes: (1) a multi-scale DCT tokenizer that compresses a 224x224 image into compact frequency-domain tokens, achieving up to 128x reduction in off-chip data volume with modest accuracy loss; (2) a LUT-based near-sensor hardware implementation that leverages fixed DCT coefficients for multiplier-free, energy- and area-efficient tokenization; and (3) a modified MIPI-based low-power communication architecture that further reduces transfer energy. FrequencyFormer serves as a drop-in replacement for standard ViT patch embedding and remains compatible with pretrained backbones across classification, detection, and segmentation tasks. The pipeline achieves 28.8 TOPS/W, reduces communication energy by 230x, and lowers total sensor-side energy by 2.22x, demonstrating frequency-domain tokenization as a scalable foundation for in-sensor ViT deployment.

Vision Foundation Models (VFMs) with Vision Transformer (ViT) backbones, such as DINOv2, have become essential for downstream tasks like object recognition and semantic segmentation. The immense computational requirements of backbones often necessitate distillation into smaller architectures for edge deployment. Feature-based knowledge distillation (KD) often suffers from the teacher-student gap; the student struggles to imitate teacher's complex feature map due to its limited capacity. To mitigate this bottleneck, we propose LEAP: Layer-skipping Efficiency via Adaptive Progression, a training curriculum for ViT feature-based knowledge distillation. By utilizing the teacher's intermediate feature maps as a sequence of progressively more difficult targets, our curriculum allows the student to build a foundational representation before tackling higher-level abstractions. Our results demonstrate that this paradigm significantly accelerates convergence through adaptive difficulty selection across various student model sizes and dataset scales. With our curriculum, the LEAP-distilled ViT-S achieves 90.1% accuracy on ImageNet-100, a +12.24% improvement compared with baseline. On ImageNet-1K, LEAP achieves +3.84% and +7.75% improvement for the instance retrieval task on the Oxford and Paris datasets, respectively. Furthermore, the curriculum enables 25.1% savings in training FLOPs and 21% savings in training time on ImageNet-100 by implementing early-stopping for teacher inference during the initial stages of training. Code is available at https://github.com/KevinZ0217/LEAP

Despite the widespread adoption of Vision Transformers (ViTs) and their success across numerous computer vision applications, the fundamental understanding of their dimensional and representational geometry remains relatively underexplored. To address this gap, we introduce Transformer Geometry Observatory (TGO), a systematic framework of experiments and analysis pipelines designed to investigate the representational geometry and dynamics of Vision Transformers. TGO-I, the first installment of the framework, focuses on the spectral geometry of ViT representations. Using a ViT-Small/16 model trained on ImageNet-100, we analyze Effective Rank, Stable Rank, Participation Ratio, Spectral Entropy, Spectral Flatness, Spectral Anisotropy, covariance structure, eigenspectra, and singular value spectra throughout training. Our results reveal a consistent increase in dimensional utilization, accompanied by decreasing anisotropy, increasing spectral entropy, increasing participation ratio, and progressively flatter eigenspectra. Contrary to the common intuition that training should concentrate information into a small number of dominant directions, we observe a progressive redistribution of variance across representational dimensions. This phenomenon is particularly pronounced in the final CLS token representation, which exhibits the highest effective dimensionality and lowest anisotropy within the network.

Face Presentation Attack Detection (PAD) systems constitute a critical security layer in biometric authentication; however, existing approaches exhibit systematic performance disparities across demographic groups, disproportionately affecting individuals with darker skin tones. This paper presents a comparative empirical investigation of whether Vision Transformer architectures reduce demographic bias in face PAD systems relative to convolutional baselines. Experiments are conducted on the CASIA-SURF Cross-Ethnicity Face Anti-Spoofing (CeFA) dataset. Three architectures are evaluated: a Multimodal ViT-Tiny trained from scratch, a ResNet18 CNN baseline, and a pretrained DeiT-S fine-tuned on CeFA across African, East Asian, and zero-shot Central Asian demographic groups. DeiT-S achieves the highest overall accuracy of 97.27% and the lowest EER of 0.86%, outperforming ResNet18 at 90.15% accuracy. In terms of fairness, DeiT-S reduces the inter-ethnic ACER gap between African and East Asian subjects to 0.13%, compared to 0.75% reported in an LBP-based work [6], representing an 83% reduction. Most notably, while ResNet18 records a BPCER of 10.44% on zero-shot Central Asian subjects, DeiT-S maintains 2.89% on the same unseen group, demonstrating a 3.6x generalization advantage. These results suggest that pretrained Vision Transformers achieve superior PAD accuracy, produce smaller demographic performance gaps, and generalize more equitably across unseen demographic groups, indicating that cross-demographic fairness in PAD may partly be influenced by architectural design.

Coastal algal bloom monitoring requires frequent, spatially detailed, and globally consistent observations, provided by Landsat-8/9 and Sentinel-2 A/B/C. Together, these missions offer over a decade of medium-resolution multispectral imagery with near-global coverage every 2-3 days, enabling the detection of fragmented bloom structures not resolvable by coarse ocean-color sensors. However, their use in aquatic environments remains challenging due to limited spectral coverage and a lack of harmonized reflectance products. As an alternative to traditional bio-optical methods, deep learning-based image classification offers a data-driven approach that can overcome many of these limitations. This study presents the first successful implementation of vision transformer-based coastal algal bloom mapping using 30-m Landsat-Sentinel-2 images. A globally distributed bloom patch dataset was generated across bloom-prone coastal hotspots worldwide. Four transformer-based architectures were compared against a standard convolutional baseline for fine-scale bloom detection, and assessed under different optical water types and atmospheric and surface conditions. All deep learning models showed strong capabilities in detecting floating bloom areas, with omission and commission errors of 8-65%. Under cloud and glint stress in a time series, the Swin Transformer outperformed traditional spectral-index approaches, which produced widespread false positives, effectively avoiding cloud- and glint-affected pixels. Comparisons with MODIS-derived products further highlighted the benefits of higher spatial resolution in detecting fragmented and irregularly affected blooms. Our findings support deep learning as a reliable tool for medium-resolution, consistent monitoring of floating algal blooms in dynamic coastal environments.

This study proposes an overall deep learning architecture for multi-class classification of plant diseases from high-resolution leaf imagery, with a particular interest in investigating the behavior of ResNet-50 and a hybrid ResNet + Vision Transformer (ViT) design. A specially gathered image database with 15,200 training images and 3,800 validation images spanning 38 classes across multiple crops, including tomato, apple, grape etc. were subjected to preprocessing steps such as resizing, normalization, and data augmentation to enhance model robustness. Multiple architectures, including ResNet-50, MobileNetV2, and EfficientNet-B0, were trained and compared with the hybrid ResNet + ViT model. All models were fine-tuned using the AdamW optimizer and cross-entropy loss, with early stopping applied to prevent overfitting and ensure generalization. Furthermore, interpretability techniques such as Grad-CAM and saliency maps were implemented to indicate disease-relevant regions, while segmentation-based analysis was performed to identify the affected parts of a leaf. For every one of the considered architectures, ResNet-50 led to the highest accuracy of 98.74%, whereas the hybrid ResNet + ViT model achieved a competitive accuracy of 98.58%, showing that the hybrid architectures were effective in capturing both local and overall information. The experimental results showcase the promise of transformer-based models to achieve highly accurate, interpretable, and computationally efficient computer-based multi-class multi-disease classification systems, providing helpful assistance for cultivation management practices as well as for precision farming.

Vision Transformers (ViTs) have become a dominant architecture for visual representation learning, providing exceptionally strong and broadly reusable backbone features. However, ViTs are commonly operated on relatively small patch-token grids due to the quadratic cost of global self-attention, which creates a persistent bottleneck for dense prediction tasks such as semantic segmentation and depth estimation. This has motivated the development of task-agnostic feature upsamplers. While recent state-of-the-art methods produce visually sharp dense representations, their reliance on shallow image encoders for guided upsampling can introduce feature leakage, fragmentation, and blur. We introduce ViT-Up, an implicit feature upsampling framework that replaces external image guidance with layer-wise query construction from intermediate ViT hidden states. This enables feature prediction at arbitrary continuous image coordinates while preserving alignment with the backbone feature space. Experiments demonstrate that ViT-Up consistently outperforms state-of-the-art image-guided upsamplers across dense prediction and semantic correspondence. On DINOv3-S+, ViT-Up improves over prior methods by up to +2.07 mIoU on Cityscapes and +4.17 PCK@0.10 on SPair-71k. With the larger DINOv3-B backbone, these gains increase to +3.36 mIoU and +8.09 PCK@0.10, demonstrating that ViT-Up scales favorably with backbone capacity.

Recent advances in Vision Transformers (ViTs) for face recognition (FR) have moved beyond the standard CLS-token paradigm. In this paradigm, a special classification token (CLS) is prepended to the patch embeddings and used as a representation of the input for downstream tasks. An alternative approach, Concatenated Patch Embeddings (CPE), instead leverages all patch tokens by concatenating them into a single vector, which is then projected into a compact face representation. CPE has been shown to improve recognition performance in comparison to CLS-based ones, but our qualitative analysis of attention maps showed the presence of artifacts that limit their interpretability. To address this issue, we incorporate register tokens, learnable tokens concatenated to the initial patch embeddings, and processed jointly through the ViT encoder blocks. This mechanism has been shown to produce more structured and interpretable attention maps compared to baseline ViT. We empirically demonstrate that these artifacts consistently appear across various ViT backbones, including small and large models, and that introducing register tokens effectively mitigates them. Adding four or eight registers significantly enhances interpretability, with eight registers providing the highest verification accuracies and smoothest attention structures. Our resulting model, ViT-8R, corresponds to a CPE-based ViT-B architecture augmented with eight register tokens achieves state-of-the-art performance among ViT-based FR models on large-scale IJB-B and IJB-C benchmarks. Also, ViT-8R produces substantially clearer attention maps compared with the baseline model, which offer deeper insight into the model's attention behavior (https://github.com/TaharChettaoui/ViT-FR-Registers)

Vision Transformers (ViTs) have gained significant attention in computer vision and shown strong potential for face recognition (FR). However, their high computational cost makes deployment on resource-constrained devices challenging, motivating the need for methods that balance efficiency and accuracy. In this work, we investigate early exiting in pretrained ViTs as a simple yet effective training-free strategy for efficient FR inference. Leveraging the uniform feature dimensionality across transformer encoder blocks, we introduce ViT-FREE, a multi-exit framework that enables face verification directly from intermediate representations without modifying or retraining the backbone model, and thus, reducing inference cost. Empirically, we show that patch embeddings and attention maps evolve progressively across depth, exhibiting high similarity between consecutive ViT blocks and increasing alignment with the final representation. This indicates gradual feature refinement and attention convergence, suggesting that intermediate layers already provide stable and discriminative representations suitable for early exiting. Through extensive experiments on multiple FR benchmarks, we systematically analyze the accuracy-efficiency trade-off across exit depths. Our results demonstrate that later exits achieve a highly favorable balance, with exiting at layer 10 yielding up to a 20% speedup while incurring only a 1.5 drop in verification performance on benchmarks such as IJB-C. Also, we propose ViT-FREE_FT, a lightweight exit-specific fine-tuning strategy that adapts only the projection layers using a small synthetic dataset while keeping the transformer backbone frozen. This approach improves the performance of shallow exits while preserving the efficiency benefits and leaving deeper exits largely unaffected.

Vision sensors provide a lightweight solution for spacecraft proximity operations, but monocular spacecraft 6D pose estimation remains difficult under illumination variation, specular reflection, shadowing, weak texture, and background interference. These factors make local visual evidence spatially unreliable and can destabilize pose regression. This article proposes a Precision-Aware Illumination-Disentangled Vision Transformer (PAID-ViT) for robust spacecraft pose estimation.The proposed model separates pose-relevant structure tokens from illumination-sensitive appearance tokens, estimates patch reliability before pose aggregation, and uses foreground mask supervision to preserve silhouette cues. A parameter-free geometric recovery module converts normalized crop coordinates, log-depth, and a continuous 6D rotation representation into camera-frame rotation and translation. Experiments on SPEED+ V2, the SPEED+ validation/lightbox/sunlamp evaluation configuration used in this study, suggest that PAID-ViT reduces translation error and improves robustness in the challenging sunlamp domain, while ablation studies support the complementary roles of illumination disentanglement, reliability-aware token aggregation, mask supervision, and training-side regularization.

Though Vision Transformers (ViTs) have become the dominant backbone in many computer vision tasks, due to permutation equivariance, their attention mechanism lacks explicit spatial inductive biases. This become particularly important in two settings: when model capacity is small or training data is limited. Inspired by the attention masking strategies in Linear Transformers and the scanning patterns of Vision SSMs, we introduce VIOLIN, a lightweight masked attention mechanism that encodes spatial structure within attention via Space Filling Curves (SFCs) with less than 0.0015% extra parameters and negligible computational overhead. VIOLIN scans the image using multiple SFCs to construct curve-specific decay masks, which are then combined and multiplied with the attention matrix. Across a wide range of evaluations, VIOLIN consistently improves performance. In limited data regimes such as fine-tuning on VTAB-1K, it boosts accuracy across all task groups and by up to 8.7% on the tasks where spatial information is essential. It can be combined with parameter-efficient fine-tuning methods such as LoRA to further increase the performance. Beyond fine-tuning, VIOLIN improves various small scale ViT architectures (e.g., DeiT, DINO) during pretraining on ImageNet-1K. Additionally, on pixel-level CIFAR-100 training, a task that is highly dependent on location information, VIOLIN increases accuracy by up to 7.2%. Overall, VIOLIN provides a computationally efficient yet effective way to inject spatial inductive bias into ViTs, especially benefiting small models and limited data settings.

Image captioning is a challenging and significant task that aims to generate coherent and semantically meaningful textual descriptions for given images. To accomplish this task, it requires a deep understanding of visual content along with the ability to express that understanding in natural language. Despite remarkable progress with transformer-based architectures, existing approaches often suffer from limitations, such as a lack of rich local feature representations and the high computational cost of quadratic self-attention. The proposed model focuses on improving computational efficiency by restructuring the vision transformer architecture. In designing this approach, the standard self-attention mechanism in Vision Transformers is replaced with a probabilistic transformer approach based on a Gaussian Mixture Model (GMM), a soft-clustering technique. Instead of computing pairwise attention among all image patches, the model groups similar patches into a fixed number of clusters using an Expectation-Maximization (EM) algorithm. This clustering-based mechanism reduces the computational complexity from quadratic O(n^2) to linear O(nK), where K << n. The autoregressive GPT-based decoder is used for caption generation. The model is evaluated on the Flickr 30K dataset, demonstrating competitive and significant improvement over existing works.

Temporomandibular joint osteoarthritis (TMJ OA) is a prevalent degenerative condition whose osseous changes are often subtle on cone-beam CT (CBCT), making automated detection challenging. We study how well the DINO family of self-supervised vision transformers -- DINOv1, DINOv2, DINOv2+reg, and RAD-DINO (a radiology-pretrained variant) -- transfers to CBCT, asking how much backbone adaptation is needed and of what kind. We propose a simple slice-based pipeline using Vision Transformer (ViT) backbones: axial CBCT slices are encoded per-slice by a frozen or partially adapted ViT and aggregated via attention-based multiple instance learning (MIL) for patient-level binary OA/Normal classification. Through systematic ablation across unfreezing strategies and aggregation designs on a multi-source CBCT dataset, we find that partial unfreezing of the final two transformer blocks is the decisive factor, improving AUC from 0.671 (fully frozen DINOv2) to 0.902. This outperforms DINOv1 (0.867), DINOv2+reg (0.774), and a supervised ImageNet ViT-B/16 baseline (0.843). Our results provide practical guidance for adapting DINO-family foundation models in low-data medical imaging settings, showing that adaptation strategy is a stronger driver of performance than backbone choice alone.

Vision Transformers (ViTs) achieve strong performance but suffer from high computational costs due to quadratic self-attention complexity. Although token reduction techniques such as pruning and merging mitigate this, they typically overlook how representations evolve across network depth. We propose RAPID, a depth-aware token reduction framework that adapts reduction strategies to the layer-wise characteristics of token representations. The primary methodological contribution is a bifurcated strategy: in shallow-to-middle layers, RAPID employs a redundancy-similarity aware pruning metric to eliminate over-represented local patterns. As features transition to global semantic concepts in deeper layers, the framework shifts to an importance-similarity aware merging mechanism. This stage leverages classification (CLS) token attention weights to protect semantically critical tokens while fusing less important but similar neighbors. Empirical validation on ImageNet-1K using ViT and DeiT architectures demonstrates that RAPID establishes a superior accuracy-compression Pareto frontier compared to plug-and-play baselines such as ToMe and ToFu. RAPID is particularly robust in aggressive compression regimes, achieving up to 4.29% higher accuracy than ToMe at extreme reduction rates. Our framework provides a training-free template for optimizing vision models by aligning reduction strategies with hierarchical feature evolution.

Vision Transformers operate on fixed patch grids, which can introduce phase-dependent instability for dense prediction: changing the patch partition can change the token evidence available to a pixel, especially near boundaries. We formalize patch-grid phase as a nuisance variable and propose Phase Marginalization, a post-hoc marginalization method that evaluates structured patch-grid phases, inverse-aligns dense outputs, and aggregates them in the original image coordinate system. The central variant, Uniform Phase Marginalization with K = 4, is training-free and improves over the canonical K = 1 baseline across measured segmentation, depth, and local matching settings. In a controlled Cityscapes experiment, Uniform Phase Marginalization provides a modest compute-matched advantage over generic shift-based four-forward test-time augmentation (TTA) (+0.31 mean Intersection-over-Union over the strongest tested generic row). A scaling study further shows that K = 4 is a practical cost-accuracy trade-off: K = 8 is essentially unchanged and K = 16 adds little accuracy at much higher latency. These results position patch-grid phase as a measurable nuisance variable and Phase Marginalization as a simple diagnostic and post-hoc marginalization baseline for dense ViT prediction.

Despite high accuracy, Vision Transformer (ViT) predictions can be driven by spurious cues, raising the need to understand their inner workings before safe deployment. Sparse autoencoders (SAEs) provide a promising lens for decomposing model representations into human-interpretable concepts, yet adapting SAE-based interpretation to ViTs remains challenging due to limited control over concept coverage and subjective, non-scalable feature interpretation. To fill the gaps, motivated by neuroscience-inspired principles, we propose ViSAE, a mechanistic interpretability toolbox for understanding ViT inner workings through concept circuits. ViSAE consists of three components: (1) A probing suite with 64K images and a 16K visually grounded concept vocabulary, improving concept coverage efficiency by 20x over ImageNet and interpretation accuracy by 28.7% over existing concept sets. (2) Top-down concept reading and Bottom-up circuit tracing algorithms that automatically recover ViT inner workings via concept circuits. (3) Applications for auditing and steering ViT behavior. Through concept editing, ViSAE improves the worst-group accuracy on WaterBirds by 48.2%, outperforming existing methods by 23.8%. Our data and code: https://github.com/deep-real/ViSAE.

In remote sensing object detection, Convolutional Neural Networks (CNNs) excel at capturing local details while Vision Transformers (ViTs) are better at global context modeling. However, existing detectors typically rely on a single fixed backbone or a manually designed hybrid architecture, and thus fail to adaptively exploit these complementary strengths across inputs of diverse complexity. To address this limitation, we propose Backbone Module Composition via Reinforcement Learning (BMCR). BMCR dynamically assembles input-adaptive inference paths from reusable modules decomposed from off-the-shelf CNN and ViT backbones. To enable such cross-family composition, we first construct an extensible module toolbox. Specifically, we decompose representative CNN and ViT backbones into reusable functional modules and encapsulate each module with explicit structural, semantic, and computational metadata for compatibility-aware assembly. To bridge the gap between grid-based CNN features and token-based ViT representations, we design a lightweight Optimal Transport (OT) based transition interface that ensures distribution-aware alignment while respecting spatial consistency. The backbone composition process is then formulated as a sequential decision problem, in which a policy network progressively selects task-relevant modules according to intermediate multi-scale observations. To stabilize the joint optimization of reusable modules and the routing policy, we further develop an Adaptive Module Cooperative Optimization (AMCO) strategy that coordinates module updating, routing exploration, and reward assignment during training. On DOTA-v1.0, DOTA-v1.5 and DIOR-R, BMCR achieves 79.31%, 73.41% and 71.86% mAP, respectively, surpassing strong static and dynamic baselines by up to 2.5 points while maintaining competitive efficiency.

Data-Free Quantization (DFQ) addresses data security concerns by synthesizing samples, without accessing real data. It has garnered increasing attention in the context of Vision Transformers (ViTs), owing to the superiority of the self-attention mechanism compared to classical convolutional operation. However, previous DFQ arts for ViTs often suffer from a distribution mismatch between synthetic samples and input distribution expected by quantized models Q, resulting in the suboptimal performance. In this paper, we propose a novel Masked Attention Alignment approach for Data-Free Quantization of ViTs, named MaskAQ, revealing that: 1) the semantics in the self-attention mechanism is predominantly localized to a sparse subset of patches, called informative regions; 2) the informative regions dominate the mutual information between synthetic samples and Q's outputs. To these ends, we incorporate differential entropy maximum over patch similarity of synthetic samples, to decouple informative regions from noisy background. To couple with varied Q, the informative regions are selected to align full-precision models with Q via a masked attention alignment objective, thus yielding high-quality synthetic samples. Furthermore, a periodic sample refreshing strategy comes up to endow MaskAQ with the capacity to continually adapt to the evolving state of Q throughout the training process, to preserve desirable mutual information with synthetic samples. Extensive experiments verify the merits of MaskAQ over state-of-the-art approaches across multiple backbones and downstream tasks. Our code is available at https://github.com/hfutqian/MaskAQ.

Representations of the world, arguably, contain information about features (e.g. something is blue, something is a circle) but also information about which features are part of the same object (e.g. the circle is blue), which we call binding information. Any system with the ability to understand scenes with multiple objects must be able to solve the binding problem: it needs to know which features belong together. However, despite work showing that Vision Transformers (ViTs) know which patches belong together, it is not known whether current deep learning models learn to exhibit binding information, i.e., for features. We may believe that there is not much binding information, after all misattributing features to wrong objects is a common failure of ViT-based architectures, especially in scenes with objects sharing features. Here we formalize the binding problem with an information-theoretic approach, and introduce a probing method to measure binding information in model representations. We perform experiments on ViTs, measuring binding from different components of the architecture, such as the image summary token [CLS] or the spatial tokens. We use datasets with different binding challenges, such as feature sharing, occlusion, and natural features, while comparing the performance of several pre-trained ViTs. Overall, our research demonstrates binding as a key ingredient to strong visual recognition and reasoning.

Robust visual classification often depends on localizing the main foreground objects in an image while ignoring contextual distractors. Surprisingly, we find that the attention maps of smaller self-supervised ViTs localize foreground objects better than those of larger ViTs. However, we still need large ViTs, because they extract richer representations from each patch. To get the best of both worlds, good localization and rich representations, we propose $A^2$, a simple method that leverages this inverse scaling finding by decoupling where to look (a small attention model) from what to extract (a large embedding model): we crop around the attention peaks of a small model and embed the crops with a larger model. $A^2$ uses entirely pretrained features, requires no group labels, and does not require per-dataset attention or backbone training. Across 5 benchmarks, $A^2$ is competitive with backbone-matched loss-level methods like DFR, and outperforms end-to-end attention training under stronger distribution shifts.

Accurate brain tumor classification from magnetic resonance imaging (MRI) is a key requirement for early diagnosis and clinical decision-making. Vision Transformers (ViTs) have shown strong performance in medical image analysis by learning global contextual representations, but they often fail to capture intrinsic structural and topological patterns present in tumor regions. To address this limitation, we propose a fusion framework that combines Topological Data Analysis (TDA) features with pretrained Vision Transformer representations for four-class brain tumor classification. In the proposed method, TDA is used to extract complementary topological descriptors that capture geometric structure, connectivity, and shape information from MRI images. In parallel, a pretrained ViT model learns high-level semantic representations from the same images. These two feature spaces are then fused to form a unified and more discriminative representation for classification. The model is evaluated on the BRISC2025 dataset, which contains four brain tumor classes: glioma, meningioma, pituitary tumor, and non-tumor cases. Experimental results show that combining topological and transformer-based features significantly improves performance compared to using either approach alone. The proposed TDA-ViT fusion model achieves an accuracy of 99.10%, precision of 99.27%, recall of 99.15%, F1-score of 99.21%, and an AUC of 99.98%. It also outperforms several state-of-the-art models, including ResNet50, ResNet101, EfficientNetB2, and standalone Vision Transformers. These results demonstrate that topological features provide valuable complementary information that enhances deep representation learning, leading to a robust and highly accurate framework for automated brain tumor classification.

Compact Vision Transformers are attractive for medical imaging in low-data and resource-constrained settings, but most existing variants assume that Euclidean latent geometry is sufficient for organizing image representations. We introduce hZACH-ViT, a family of curved-geometry extensions of ZACH-ViT, a compact zero-token Vision Transformer that removes positional embeddings and the class token and relies on global average pooling over patch representations. To isolate the role of geometry, we preserve the verified ZACH-ViT backbone and modify only the final representation space and prototype-based classifier head, enabling a controlled comparison between Euclidean, hyperbolic, and spherical latent geometries. We evaluate Poincaré, Klein, and spherical hZACH-ViT heads on seven MedMNIST datasets under an identical few-shot protocol with 50 samples per class and five random seeds. The completed benchmark contains 770 training runs spanning seven datasets, three non-Euclidean geometries, seven curvature magnitudes, and a Euclidean baseline. Across all seven datasets, the best non-Euclidean hZACH-ViT configuration improves over Euclidean ZACH-ViT, with an average gain of +0.021 in the dataset-specific primary metric and the largest improvement on OCTMNIST (+0.055 MacroF1). Fixed low-curvature configurations retain positive gains on the majority of datasets, and low curvature values (c = 0.1 or 0.2) account for six of the seven dataset-level winners. Rather than identifying a universally optimal manifold, our results establish geometry and curvature as dataset-dependent model-selection variables, with fixed low-curvature analyses confirming that gains persist beyond exhaustive per-dataset tuning.

With the growth of Vision Transformers in safety-critical domains like autonomous systems and medical imaging, ensuring their reliability against soft errors is paramount. While ViTs offer state-of-the-art accuracy, their massive parameter counts render exhaustive fault injection campaigns infeasible. To bridge this gap, a statistical fault injection framework is presented, leveraging finite-population sampling theory to provide formal reliability guarantees. It is demonstrated that failure rates are bounded within a 1% margin at 99% confidence using only a few thousand samples, regardless of model scale. This methodology achieves up to a 10,700 times reduction in experimental cost compared to exhaustive approaches, while preserving the ability to localize vulnerabilities across architectural components. Through extensive evaluation of different architectures like ViT-Tiny and ViT-Small, a highly non-uniform reliability landscape is uncovered. It is shown that while only 3% of FP32 bit-flips result in failure, the vast majority of these events lead to catastrophic accuracy collapse. Specific vulnerabilities are localized to normalization layers and critical exponent bits within the IEEE-754 format, providing a mathematical foundation and actionable insights for the design of hardened, edge-deployed ViT architectures.

The widespread use of image classification models in high-risk, real-world situations necessitates making these models robust to slight disturbances or perturbations, such as blurring or sharpening, in the input images. While vision transformers (ViTs) play an integral role in many modern-day multi-modal models like Vision-Language-Models (VLMs) and Vision-Language-Action (VLA) models, they have received a lack of attention in the setting of robustness. In this work, we analyze the effects of adversarial fine-tuning, a popular method for improving model robustness to image perturbations, on a ViT's performance on perturbed and regular images through a mechanistic lens. We adversarially train a ViT on low-frequency and high-frequency image corruptions, and attempt to explain changes in downstream model performance through an examination of the model's attention mechanisms, internal representations, and knowledge evolution. Overall, our results suggest that, while fine-tuning on inputs with common corruptions improves model performance and certainty on new instances of corrupted data, these improvements do not transfer to other classes of corruptions not seen in the training. Additionally, despite observing changes in visual attention and knowledge evolution across layers, we found that adversarial training did not lead to fundamental changes in the sparse representations learned by ViTs.

Positional embeddings (PEs) in Vision Transformers (ViTs) are known to impact performance and robustness, but their role in shaping internal spatial representations is not well understood. In this work, we study how different forms of PEs influence the representational geometry of ViTs and how these changes relate to robustness under content-disrupting distribution shifts. We introduce a metric, the Spatial Similarity Distance Correlation (SSDC), to quantify spatial structure in token representations. Using this metric, we show that ViTs trained without PEs still develop non-trivial spatial structure, but this structure is driven by visual content and collapses under token permutation. In contrast, we find that all PEs considered (learned absolute, sinusoidal, and rotary) are associated with a consistent shift toward an index-anchored spatial organization. Representations in these models remain stable under perturbations that disrupt content, and exhibit substantially improved robustness to such distributional shifts. We further show that while different PEs produce distinct depth-wise trajectories of spatial structure, their robustness properties are largely similar (with secondary variation across encoding schemes), suggesting that robustness appears to depend on the presence of a stable positional reference frame more than it depends on the specific encoding mechanism. These results offer a geometric account of how positional encodings shape internal representations, with implications for the principled design of future encoding schemes.

Histological scoring is essential for diagnosing Non-Alcoholic Fatty Liver Disease (NAFLD), yet its automation remains challenging due to the high annotation cost and negative transfer among the strongly correlated NAFLD Activity Score (NAS) indicators in multi-task learning. To address this issue, we propose a subspace-decoupled multi-task Vision Transformer (ViT) that integrates lightweight task-specific Adapters with orthogonality-based constraints. This design constructs independent feature subspaces for steatosis, ballooning, and inflammation, effectively reducing task interference while retaining shared representations. We further construct a curated multi-task mouse NAFLD histology dataset with expert annotations for all NAS components. Experimental results demonstrate that the proposed method improves multi-task stability and generalization with substantially reduced computational cost compared to training separate single-task models. The code and the curated dataset have been prepared and will be made publicly available upon acceptance to support reproducibility.

This study aims to enhance maritime security by using advanced Artificial Intelligence (AI) and Computer Vision (CV) techniques. For this purpose, it was designed and assessed intelligent object detection systems that can detect the presence of ships on the sea surface under different real-time environments. To achieve this goal, a maritime image dataset with 6,468 images was used, covering different weather conditions like cloudy, foggy, rainy, and sunny environments. Six deep learning architectures were evaluated, including a base Convolutional Neural Network (CNN) model, four transfer learning models (Xception, VGG16, MobileNetV2, and EfficientNetV2L), and a Vision Transformer (ViT) model. The models were compared using multiple performance indicators, including accuracy, Type I and Type II errors, model size, and video processing time. The results show that model performance varies depending on computational constraints and deployment conditions. While lightweight architectures are suitable for resource-limited devices, the ViT achieved the best overall performance, reaching 100% accuracy with the lowest error rates and the fastest video processing time. The findings highlight the potential of AI-driven computer vision systems for maritime surveillance, border protection, and autonomous navigation.

Self-supervised learning (SSL) has produced a diverse landscape of vision transformers (ViTs) whose pretrained representations support a wide range of downstream tasks. Towards a better understanding of these models, a body of work has assessed the mechanics of their self-attention as well as the types of information captured across their representations, revealing, for example, stark differences between models trained with contrastive learning (CL) and masked image modeling (MIM). However, the total of these advances on model understanding has to date not yet fully permeated a larger community, where, e.g., insights that are specific to CL models are still at times generalized to MIM models. To make model understanding straightforward and intuitive for a broad community, we propose a simple and easily interpretable visualization protocol. Our protocol is based on visualizing unsupervised semantic segmentation results, yet by no means do we focus on top segmentation performance. Instead, our protocol allows us to easily convey model behavior that consistently emerges across images. Benchmarked on a diverse set of SSL models across layers and representations, our protocol allows us to gain novel insights into distinct positional biases and scaling behaviors, including, e.g., strong boundary artifacts in DINOv3-Large model tokens. These novel insights come on top of more easily conveying a range of previous findings. Our protocol further allows us to clearly visually convey and distinguish between positional effects and the closely related but distinct locality bias, the latter being much more extensively studied in the literature so far. Our protocol is publicly available, serving to catalyze further model understanding for a broad community.