Attention
-
CondenseVLA: Learnable History Condensation for Efficient Multi-Frame VLA
2026-10-20 15:47:30Feiyang Hong, Yujie Wei, Bo Zhao, Xiu Su, Hongxun Yao, Shuo YangAbstract
Robot manipulation often depends on short-term cues - such as brief contacts, temporary occlusions, and intermediate object states - that are ambiguous in a single frame and can only be resolved by integrating observations over time. However, injecting historical information solely by enlarging the history window is computationally costly and frequently yields limited gains due to substantial redundancy and task-agnostic, inflexible sampling of observations. We propose CondenseVLA, a multi-frame VLA policy with a learnable decoder-side history condensation module for using short-term visual context in action decoding. Given a sequence of cached per-frame features from the VLM backbone, CondenseVLA distills observation history into a fixed set of slot tokens using learnable queries and multi-head cross-attention. Simple auxiliary regularization penalizes overlap among slot attention maps and decorrelates slot embeddings, encouraging complementary attention and reducing slot redundancy. Across the evaluated tasks, CondenseVLA improves Simpler-WidowX success from 58.7% to 69.8% (+11.1 percentage points), achieves 97.2% on LIBERO, and reaches 52.8% on real-world tasks. It achieves 21.6% success on memory-dependent RMBench tasks. These results support selective condensation of bounded recent visual evidence for action decoding, while harder multi-observation tasks expose the limits of a finite history window. Project page: this https URL
Abstract (translated)
URL
https://arxiv.org/abs/2610.25927
PDF
-
RACE: RoPE-Aligned Attention Compression Based on Query-Key Activation Energies
2026-10-20 15:34:02Irene Tenison, Abdulrahman Alabdulkareem, Daniel Karl I. Weidele, Nima Dehmamy, Dan GutfreundAbstract
Head-dimension compression can reduce the memory required by the key-value (KV) cache, but existing methods often reconstruct full-dimensional representations during inference, adding significant decoding FLOPs overhead. We formalize the design considerations for converting pretrained attention into a natively reduced-dimensional form that reduces both KV-cache storage and attention computation using a calibration set. Specifically, we consider reductions that (1) apply a shared projection to queries and keys so that they remain in a common reduced feature space, (2) derive this projection jointly from query and key activations to better preserve their interaction, and (3) are compatible with the pretrained RoPE transformation. To obtain a reconstruction-free realization of this reduction, we restrict it to coordinate selection and theoretically show that RoPE compatibility then holds if and only if complete rotary pairs are retained. Building on this characterization, we introduce RACE, a post-training method that selects rotary pairs using their joint query-key activation energy. We theoretically show that this criterion minimizes a separable, calibration-dependent upper bound on post-RoPE query-key product approximation error. The resulting reductions are folded into the model weights and require neither runtime reconstruction nor specialized kernels. Across three model families and different compression levels, RACE outperforms structured pruning and remains competitive with reconstruction-based methods on language-modeling, zero-shot, and long-context tasks. At 50% KV-cache retention and 256K context, RACE reduces total inference FLOPs by 47.8%, yielding 1.79x higher throughput, 43.2% lower time-to-first-token (TTFT), and 28.1% lower allocated GPU memory in comparison to an uncompressed model.
Abstract (translated)
URL
https://arxiv.org/abs/2610.28385
PDF
-
HistoLoop: Iterative Contextual Inference with a Looped Transformer for Histology-Based Spatial Gene Expression Prediction
2026-10-20 15:05:54Ziang Liu, Jie Liu, Shuai Li, Yigui Feng, Qinglin WangAbstract
Predicting spatial gene expression from H&E images offers a scalable complement to experimental spatial transcriptomics. However, despite progress in spatial context modeling, how to construct subsequent context from spot states updated by earlier contextual interactions remains underexplored. We introduce HistoLoop, a looped Transformer that formulates prediction as iterative contextual inference. At each refinement step, HistoLoop aggregates the current spot states into context tokens at the corresponding resolution, processes them with a shared spatial Transformer, and retrieves spot-specific contextual information through cross-attention. A memory-mediated update integrates this information into persistent spot states, which are then used to construct context at the next, finer resolution. Under the HEST-Bench HVG50 evaluation protocol, HistoLoop achieves the highest macro-averaged gene-wise Pearson correlation among the compared methods across six cancer-specific tasks spanning five organs in HEST-1K. Ablation studies support the importance of recurrent state refinement and reconstructing context from updated spot states at progressively finer resolutions. Expression-derived clustering further suggests that HistoLoop better captures spatially organized multigene patterns.
Abstract (translated)
URL
https://arxiv.org/abs/2610.28339
PDF
-
AGGR-Mamba: Mamba-based Network with Attention-Guided Gated Residual Refinement for Undersampled MRI Reconstruction
2026-10-20 14:39:35Yue Zhuo, Wenbin Deng, Lingjun Liu, Xuhang Chen, Zhonghua XieAbstract
Undersampled magnetic resonance imaging (MRI) reconstruction aims to recover high-quality images from partially acquired k-space data, thereby reducing scan time. However, missing frequency-domain measurements often lead to residual aliasing artifacts, blurred edges, and loss of local structural details. In this paper, we propose AGGR-Mamba, a lightweight output-stage residual refinement method built upon a Mamba backbone for undersampled brain MRI reconstruction. The Mamba backbone first produces an initial reconstruction, and the proposed refinement modules then correct the remaining local errors and high-frequency artifacts in a controlled residual manner. Specifically, an Attention Refine Head is introduced to recalibrate local anatomical structures, edge textures, and residual-error regions through channel and spatial attention. In addition, a GDFN-style gated residual refinement branch is designed to selectively preserve informative residual features while suppressing redundant or unstable responses. By combining attention-based recalibration with gated residual selection, AGGR-Mamba improves local structure recovery without substantially disturbing the backbone reconstruction. Experiments are conducted on the IXI brain MRI dataset with T1- and T2-weighted images under acceleration factors of R=4 and R=8. The results show that AGGR-Mamba achieves competitive PSNR and SSIM performance across different reconstruction settings, demonstrating the effectiveness of the proposed output-stage refinement strategy.
Abstract (translated)
URL
https://arxiv.org/abs/2610.28291
PDF
-
CPVB-RG: Cross-Prior Visual Bridging for Prior-Aware Chest X-Ray Report Generation
2026-10-20 14:38:03Sike Xiang, Shuang Chen, Julie Cox, Amir Atapour-AbarghoueiAbstract
Radiologists interpret follow-up chest X-rays by comparing current and prior examinations to identify new, resolved or progressing findings. Most report generation systems either omit this longitudinal context or fuse prior features without distinguishing interval changes from anatomy that remains stable over time. We propose CPVB-RG, a prior-aware framework that combines current and prior radiographs with available clinical text. Its Cross-Prior Visual Bridge (CPVB) contains Progression-Guided Focus Attention (PGFA) and two Cross-Attention units. PGFA is designed to contrast progression-difference attention with shared-spatial attention, aiming to reduce invariant responses whilst retaining coarse spatial correspondence between examinations. The Cross-Attention units then connect the resulting longitudinal visual evidence with clinical semantics. CPVB-RG is trained in two stages, using prior-guided image-text alignment followed by autoregressive report generation with a lightweight language decoder. Experiments on MIMIC-CXR and IU X-Ray show consistent improvements across language and clinical metrics. On MIMIC-CXR, CPVB-RG achieves 0.2847 F1RadGraph and 0.6692 VLScore, with statistically significant gains over the closest decoder-matched prior-aware baseline. These results support the effectiveness of the proposed difference-aware and spatially guided fusion design for prior-aware chest X-ray report generation.
Abstract (translated)
URL
https://arxiv.org/abs/2610.28289
PDF
-
Improved automatic speech-based cognitive assessment using neuropsychologically calibrated diagnostic labels
2026-10-20 14:37:35Madhurananda Pahar, Caitlin Illingworth, Adam Hampshire, Peter J Hellyer, Agata J Hampshire, Daniel Blackburn, Annalena Venneri, Heidi ChristensenAbstract
Although automatic detection of cognitive impairment from conversational speech is a promising way, it becomes less accurate, as large real-world datasets often contain unreliable diagnostic labels due to self-referral and limited clinical verification, and here we investigated whether neuropsychologically calibrated diagnostic relabeling improves it. We analysed 138 hours of conversational speech from 720 participants, recalibrating unreliable diagnostic labels using attention, executive function, and memory performance via external clinical thresholds, Gaussian, and normative z-score calibration. Transformer-based classifiers were trained using both original and recalibrated labels and evaluated internally on 150 clinically verified participants and externally on Pitt, ADReSS and PROCESS-2 datasets. Memory- and MoCA-based z-score calibration produced the greatest improvements. Using our internal dataset, $F_1$ increased from $0.70$ to $0.83$ ($+13\%$) and from $0.54$ to $0.66$ ($+12\%$) for binary and multiclass classification, respectively. For external validation, the increases are $+4\%$, $+5\%$ and $+9\%$ on Pitt, ADReSS and PROCESS-2 corpus. Neuropsychologically calibrated relabeling substantially improves speech-based cognitive assessment, highlighting label quality, rather than model architecture, as a critical bottleneck in clinical AI for dementia screening.
Abstract (translated)
URL
https://arxiv.org/abs/2610.28288
PDF
-
HRoPE: Homography Relative Positional Encoding
2026-10-20 14:36:51Huachen Gao, Yang Yang, Xiaoqian Liang, Xingyu Chen, Youjia Zhang, Qi Zhang, Baorui Ma, Anpei Chen, Wei YangAbstract
Geometric Relative Positional Encoding has become the primary mechanism for injecting spatial inductive biases into multi-view transformers. Existing designs condition attention on camera-level frustums or ray-marched depth, associating cross-view tokens through the sensor geometry. Such sensor-centric formulations, however, describe only where a token is observed from, leaving the rich surface geometry of the scene region depicted by the token. In this paper, we advocate a shift toward surface-aware positional encoding that explicitly grounds patch tokens in the geometry of physical surfaces. We introduce HRoPE, which represents the local geometry of each patch as a slanted 3D plane, following the slanted support-window assumption of classical multi-view stereo. We inject token-level, plane-induced homographies into attention as relative transformations. This equips the transformer with a geometric prior for rectifying perspective distortions and aligning patch features across viewpoints. Our encoding requires no ground-truth geometry: the plane parameters can be predicted end-to-end under a view-synthesis objective alone, or derived from foundation models. Extensive experiments on RealEstate10K, CO3Dv2, and DL3DV show that our best variant of HRoPE consistently outperforms the evaluated geometric positional encodings for novel view synthesis. Extensive ablations and analysis show the effectiveness of the proposed encoding. Please refer to our project page at: this https URL
Abstract (translated)
URL
https://arxiv.org/abs/2610.28284
PDF
-
Convergent Performance, Divergent Mechanisms in LoRA-Adapted Vision Foundation Models for Brain MRI
2026-10-20 14:12:44Cecilia Diana-Albelda, Fiona Kekwick, Raghav Mehta, Aleksandra Kulbaka, Paul M. Matthews, Jesus Bescos, \'Alvaro Garc\'ia-Mart\'in, Wenjia Bai, Bernhard KainzAbstract
Adapting Vision Foundation Models (VFMs) to medical imaging via LoRA has become standard practice, yet model selection is guided almost entirely by downstream task metrics, treating backbones with similar performance as interchangeable. We investigate whether equal performance implies equal internal representational shift. We analyse four ViT-B/16 backbones, DINOv3, SigLIP2, MerMED-FM, and BiomedCLIP, pre-trained with different objectives (self-distillation vs. contrastive vision-language) and on different data domains (natural vs. medical images). We adapt each to brain tumour segmentation and evaluate on six datasets covering both segmentation and Alzheimer's disease classification. Using attention divergence, weight-change analysis, and activation patching, we find that the four models converge to similar segmentation performance (mean Dice 78-81%) through markedly different internal dynamics, yet diverge in cross-task transfer: segmentation adaptation improves Alzheimer's disease classification for DINOv3 but degrades it for the other three backbones. We reveal a dissociation between modification magnitude and causal importance: the most heavily modified layers are not the causally relevant ones, with early-to-mid layers identified as sufficient for segmentation recovery. Restricting LoRA to these layers matches or approaches full-layer segmentation while reducing classification degradation in the three affected backbones, providing a mechanistically grounded strategy that is more parameter-efficient and more robust than standard full-layer adaptation.
Abstract (translated)
URL
https://arxiv.org/abs/2610.28255
PDF
-
Local Inference of LLMs on Long Contexts: A Data Systems Perspective
2026-10-20 13:56:12Mattias D\"urrmeier, Alberto LernerAbstract
Large language models can solve complex problems across a variety of subjects, such as mathematics and coding. For such complex analytical tasks, models with a long context perform better. Serving models with long context is currently mostly done on powerful systems with multiple GPUs, which are too expensive for most users. An alternative is local inference, or serving a model on a user's own machine, which removes per-token costs and rate limits, besides keeping sensitive data private and giving users full control over the models they run. The common understanding, however, is that local LLMs need compression (e.g., quantization) to run on a single GPU. In this paper, we set out to study the limitations that prevent serving long-context models for single users, besides the capacity issue. Rather than characterizing a single model on one GPU, we analyze how different models behave across several GPU generations. This lets us uncover not just individual bottlenecks but broader trends in where inefficiencies arise. Our analysis confirms that, as expected, self-attention dominates the runtime for long-context inference. More unexpectedly, our evaluation reveals that the current structure of self-attention leaves key GPU resources significantly underutilized, and the newest GPU is often the least well served. Many of these findings are fundamentally data management problems, from paging the KV cache to improving cache utilization and synchronization in the GPU, which is why we bring this study to the Data Systems community. To engage this community, we present our findings as a series of experimental observations and formulate recommendations for future work.
Abstract (translated)
URL
https://arxiv.org/abs/2610.28223
PDF
-
Can Language Models Get 'Aphasia' From Neuronal Ablation?
2026-10-20 13:49:25Pratik Worah, Randall BatemanAbstract
Progressive neuronal loss is the hallmark of neurodegenerative disease, yet how loss of neurons shapes the pattern of cognitive decline remains poorly understood. Here we systematically ablate neurons (computational units) in language models spanning recurrent long short-term memory (LSTM) and Transformer architectures and apply limited fine-tuning as a simple proxy for compensatory plasticity. With whole-unit ablation, LSTMs degrade gradually with syntactic and semantic faculties declining in parallel across mild-to-moderate damage. By contrast, Transformers exhibit late threshold-like collapse under feedforward multi-layer perceptron (MLP) ablation and smoother sigmoid-like decline under attention-head ablation. Qualitatively, MLP-ablated models continue to make within-category semantic substitutions (semantic paraphasias) alongside rising non-responses, paralleling Wernicke-type aphasia, while attention-head ablation produces a trend toward loss of semantic specificity similar to the semantic variant of primary progressive aphasia (svPPA). These findings indicate that feedforward and attentional pathways have functionally distinct failure modes, suggesting that architecture-dependent failure profiles may offer new computational perspectives on aphasic syndromes and neurodegenerative diseases as an in silico model. Notably, early accuracy retention under Transformer attention-head ablation qualitatively parallels pre-symptomatic reserve before the 40--60% neuronal loss reported by Gomez-Isla et al. (1996) when clinical symptoms emerge in Alzheimer's disease, whereas the steadier LSTM decline parallels the symptomatic progression described by DeKosky and Scheff (1990) and Terry et al. (1991).
Abstract (translated)
URL
https://arxiv.org/abs/2610.28214
PDF
-
Offline Makes Online Better: Calibrated Cache Compression for Video Generation
Abstract
Token merging is an effective approach to reducing KV-cache memory consumption in autoregressive video generation by clustering temporally redundant tokens and retaining only a small set of representatives. Existing methods, however, typically apply a uniform merging ratio across all attention layers and heads. We observe that different layers and heads exhibit substantially different yet jointly correlated sensitivities to token merging, and that these sensitivity patterns remain consistent across prompts. Based on this observation, we introduce CALM, a Calibrated Allocation for Layer-Head Merging that uses offline calibration to improve online cache compression. Exhaustively evaluating all possible layer-head allocations is prohibitively expensive. To enable efficient calibration, we show that attention-output error strongly correlates with end-to-end video quality and can therefore serve as a lightweight proxy for allocation risk assessment. We evaluate different allocation strategies using this proxy and construct a fixed layer-head risk map offline. Given a cache budget at inference time, CALM uses dynamic programming to determine the best allocation of merging ratios. Experiments across three different autoregressive video models demonstrate that CALM consistently outperforms uniform token merging under the same memory budget, by reducing the LPIPS to original videos by around 21% with higher video-quality scores.
Abstract (translated)
URL
https://arxiv.org/abs/2610.28192
PDF
-
DocLayout-YOLO: Axis-Attentive Document Layout Detector with Mesh-Guided Synthetic Pretraining
2026-10-20 13:11:41Zhiyuan Zhao, Hengrui Kang, Guang Liang, Bin Wang, Conghui HeAbstract
Accurate and efficient document layout analysis is critical for modern document parsing systems. However, although the YOLO framework represents a leading paradigm for the real-time detector, its adaptation to heterogeneous documents remains underexplored. To address this gap, we propose DocLayout-YOLO, a document-specialized real-time detector that integrates diverse synthetic document pretraining and page-axis context modeling. From the data perspective, we introduce MeshLayout-1M, a synthetic pretraining corpus generated by the proposed Mesh-candidate BestFit algorithm, which provides large-scale and diverse document layouts for pretraining and improves adaptation to heterogeneous document scenarios. From the model perspective, we propose Page-Axis Layout Attention (PALA), which captures long-range dependencies along the horizontal and vertical page axes while remaining computationally efficient. Extensive experiments demonstrate that DocLayout-YOLO achieves state-of-the-art performance on four heterogeneous document benchmarks, obtaining 72.4%/75.5%/96.0%/73.1% mAP50:95 on D4LA, M6Doc, HJDataset, and PRIMA-LAD, respectively. Compared with existing real-time detectors, DocLayout-YOLO establishes a superior accuracy-efficiency trade-off in document layout analysis. Code, model, and data are available at this https URL.
Abstract (translated)
URL
https://arxiv.org/abs/2610.28171
PDF
-
PointForward V2: Point-Aligned Interaction for Feedforward Driving Reconstruction
2026-10-20 12:30:21Xianqi Wang, Cheng Chi, Yuechuan Pu, Hongcheng Luo, Junda Cheng, Guang Chen, Hangjun Ye, Bing Wang, Xin YangAbstract
Feedforward reconstruction of dynamic driving scenes requires integrating observations across multiple views and time steps into a coherent 3D representation. PointForward established state-of-the-art feedforward driving reconstruction by lifting image observations into sparse 3D queries, demonstrating the effectiveness of point-aligned scene representations. However, these queries are largely processed independently, limiting information exchange across space and time, while direct global attention among tens of thousands of point queries is computationally prohibitive. In this work, we present \textbf{PointForward V2}, which advances point-aligned representation to point-aligned interaction. Our key idea is to mediate query interaction through compact, scene-structured registers. For the static background, frame-wise Static Scene Registers compress point information, communicate across frames, and propagate the aggregated context back to the queries. For each dynamic instance, queries from different frames are aligned in an object-centric canonical space and interact through instance-specific Dynamic Instance Registers. We further employ 4D positional encoding that combines aligned spatial coordinates with original timestamps, enabling spatially aligned yet temporally distinguishable interactions. In this way, PointForward V2 enables scalable spatio-temporal reasoning directly among lifted 3D queries without dense global attention. Experiments on Waymo, nuScenes, and Argoverse 2 demonstrate state-of-the-art performance across large-scale driving benchmarks.
Abstract (translated)
URL
https://arxiv.org/abs/2610.28120
PDF
-
MoDDiT: Simple Token Routing in Diffusion Transformers
2026-10-20 12:05:49Jarek Liesen, Matthew T. Jackson, Benjamin Ellis, Shimon Whiteson, Jakob N. FoersterAbstract
Real-time video generation, interactive world models, and on-device image synthesis are almost a reality. Diffusion Transformers (DiT) achieve excellent sample quality, shifting the bottleneck to the throughput of the models themselves. Natural images are inherently sparse, as large portions of an image are often background, contribute little to the scene content, and are easy to denoise. DiTs fail to exploit this sparsity, as they process image tokens statically and uniformly. We propose MoDDiT, a simple architecture modification that learns to route tokens through both MLP and self-attention operations. By routing tokens through both blocks independently, MoDDiT allows for dynamic compute allocation in MLPs that varies across tokens, images, and denoising timesteps. Using MoDDiT to train image generation models from scratch significantly reduces computational costs of training and inference, processing as few as 12.5% of the tokens. By increasing sparsity and model depth, MoDDiT reduces costs and improves generation performance over unrouted baselines. The learned allocation is interpretable, revealing how computation varies with visual content and noise level. We also use MoDDiT to train video generation models, where allocation correlates with motion, and to distill a dense bidirectional video model into a sparse autoregressive student.
Abstract (translated)
URL
https://arxiv.org/abs/2610.28086
PDF
-
IECG: Weakly-Supervised Interpretability-Enhancing Foundation Model for Electrocardiogram
2026-10-20 12:03:09Huynh Dang Nguyen, Trong-Thang Pham, Anh-Thuan Do, Phuong-Nhi Dinh, Brijesh Patel, Ngan Le, Van NguyenAbstract
Electrocardiography (ECG) is a cornerstone of cardiac diagnosis, yet most machine learning methods for ECG analysis lack interpretability, limiting their clinical adoption. We present IECG, a weakly-supervised ECG foundation model that jointly performs diagnostic classification and abnormality localization without requiring manual abnormal region annotations, overcoming a long-standing bottleneck in interpretable ECG learning. Central to IECG is a weakly-supervised attention strategy that explicitly incorporates clinical priors by deriving abnormality heatmap pseudo-labels from established ECG clinical knowledge. These pseudo-labels guide the model's CLS-token attention through a KL-divergence loss, encouraging it to focus on diagnostically relevant waveform regions. During pretraining, IECG is jointly optimized with clinically guided attention supervision, ECG-text contrastive alignment, and report captioning objectives, enabling it to learn rich multimodal representations while producing localized waveform evidence. By explicitly aggregating features from clinically guided attention regions, IECG grounds its diagnostic predictions in interpretable waveform evidence without sacrificing classification performance. Extensive experiments demonstrate that IECG achieves state-of-the-art abnormal region localization while maintaining strong classification performance, advancing trustworthy foundation models for ECG analysis.
Abstract (translated)
URL
https://arxiv.org/abs/2610.28083
PDF
-
Assessing Representation Semantics Under Language Pretraining in Fine-Grained Classification
2026-10-20 11:39:21Md Raisul Kibria, S\'ebastien LafondAbstract
Understanding how supervision signals shape semantic visual representations remains challenging, particularly in fine-grained, low-data settings. Standard evaluations of vision models rely on downstream accuracy, which often fails to capture structural differences in learned representations. We propose a compact explainability-driven framework for comparing encoders trained with classification-only and caption-generation supervision on CUB-200. Using a controlled two-stage protocol in which both encoders are ultimately optimized for the same classification task, we enable a direct comparison of the resulting representations. Saliency maps are generated using attention rollout and gradient-weighted rollout and evaluated with five metrics covering faithfulness (positive and negative perturbation) and localization (keypoint emphasis, background attribution, and Earth Mover's Distance). Across random seeds and explanation methods, caption supervision yields representations that are more sensitive to the removal of salient regions, better aligned with annotated object parts, and less reliant on background regions, whereas classification supervision produces representations that are slightly more robust to the removal of low-saliency regions. An ablation study on the Oxford-IIIT Pet dataset shows consistent trends across key metrics. These results demonstrate that the proposed framework makes supervision-induced representational differences observable and quantifiable beyond downstream accuracy alone. More broadly, it provides a systematic basis for assessing learned representations and comparing alternative training and transfer-learning strategies.
Abstract (translated)
URL
https://arxiv.org/abs/2610.28051
PDF
-
Recast the Past: Stable Long Video Generation with Full Context Memory
2026-10-20 10:27:07Yukuo Ma, Cong Liu, Junke Wang, Junqi Liu, Haibin Huang, Zuxuan Wu, Xuelong LiAbstract
Long-horizon video generation requires retaining an ever-expanding history, but conditioning on its growing token sequence is costly. Existing methods restrict or compress historical context, potentially losing information needed for future generation. This work proposes ReCam, which groups local spatiotemporal blocks of historical latents into vectors and projects them directly into the high-dimensional hidden space of a Diffusion Transformer. The model learns to use the resulting memory tokens through native self-attention, reducing token count while retaining block-level access to the full history. To mitigate temporal drift during autoregressive rollout, we further introduce dual-pointer denoising with separate noise schedules for recent context and target frames. Among the evaluated methods, ReCam achieves the highest VBench-Long overall score of 80.1 and competitive static-scene revisitation performance on R2M-Bench. It also leads on all four consistency metrics for off-screen subject reappearance, including clothing consistency of 4.09 versus 3.03 for the best baseline.
Abstract (translated)
URL
https://arxiv.org/abs/2610.27986
PDF
-
From Comparison to Selection: Enhancing Cross-Scale Generalization of Pointer Networks Trained with Reinforcement Learning for Capacitated Drone Delivery Routing
2026-10-20 09:44:49Sofiene Lassoued, Devara Tertia, Stefan Lier, Andreas SchwungAbstract
Unmanned aerial vehicles (UAVs) enable fast, flexible last-mile and emergency delivery, but their routing remains challenging as the number of clients varies. Classical approaches based on mixed-integer programming and heuristics can scale poorly or require problem-specific design, motivating learning-based alternatives. Reinforcement learning is particularly promising, provided a single policy can generalize across problem sizes without retraining. Although attention-based Pointer Networks are inherently size-agnostic, their solution quality degrades under both upscaling and downscaling. Using the Capacitated Vehicle Routing Problem (CVRP) as a controlled proxy, with fixed capacity reflecting drone payload constraints, we investigate this degradation and identify three contributing factors: shifts in node density, attention dispersion, and dilution of the mean-pooled graph context. We systematically compare architecture-wide, encoder-side, and decoder-side modifications to address these limitations and propose a learned, permutation-invariant graph-context aggregation mechanism, implemented as a weighted average and as Set2Set. Results reveal distinct requirements for transfer in each direction: entropy-based attention scaling reduces the CVRP20-to-CVRP100 performance gap by 1.41 percentage points, while the proposed aggregation methods achieve the largest improvement on the more challenging CVRP100-to-CVRP20 transfer, approximately 0.9 percentage points. Exploiting this complementarity, we introduce an approach selector that chooses the best model for each training-target size pair. The selector reduces cost relative to the reference model across all nine seen train-test pairs and eight of nine pairs involving unseen sizes (CVRP35, CVRP75, and CVRP200), achieving improvements of up to 3.2 percent.
Abstract (translated)
URL
https://arxiv.org/abs/2610.27936
PDF
-
Physical-Variable-Informed Deep Learning for Robust Diagnosis in Industrial Fluid Transport Systems
2026-10-20 09:12:32Yu Sha, Junqi Tao, Dixin Zhou, Horst Stoecker, Domagoj Vnucec, Andreas Widl, Kai ZhouAbstract
Industrial fluid transport systems are monitored through heterogeneous sensing streams, but robust fault diagnosis remains difficult because response signals vary with operating conditions and differ from physical variables in sampling rate, observation duration and data granularity. To address these challenges, we propose a physical-variable-informed deep learning framework (PhyNet) for robust fault diagnosis under real-world sampling heterogeneity. PhyNet treats physical variables as global operating constraints and response signals as local diagnostic evidence. Its physical-to-response cross-attention mechanism connects sequence-level physical-variable tokens with a learnable global response token, enabling interaction without strict temporal alignment. Across four benchmarks, including three real-world cavitation datasets and one public two-phase-flow dataset, PhyNet achieves the best performance across all evaluated metrics. Class-wise analyses, together with attention and masking analyses, show improved discrimination of ambiguous states and indicate that physical variables meaningfully constrain local response interpretation. These findings demonstrate a deployment-oriented approach to robust diagnosis in heterogeneous industrial environments.
Abstract (translated)
URL
https://arxiv.org/abs/2610.27891
PDF
-
Multi-View Latent Action Pretraining for Cross-Embodiment Robot Learning from Videos
2026-10-20 07:55:43Oliver Hausd\"orfer, Felix Minzenmay, Mani Majd, Maximilian P. Christof, Shiyuan Zhang, Angela P. SchoelligAbstract
Action-free videos are an appealing substitute for costly teleoperated data. Latent action models (LAMs) turn videos into a supervision signal by inferring abstract actions between frames. Existing LAMs observe a single camera view, which resolves 6D robot actions only ambiguously missing out-of-plane movements and potential occlusions. We propose a LAM that ingests synchronized multi-camera views using cross-attention to resolve depth and occlusions before compressing them into latent actions. The latent actions are fed into a policy in which a latent expert and an action expert are coupled by sequential flow matching around a frozen VLM backbone. We further evaluate in strict cross-embodiment settings where task-specific demonstrations come from other embodiments. We show that dual-view LAM pretraining improves the decodability of ground-truth 6D actions by 17% compared to single-view, while scaling single-view data by two orders of magnitude is not sufficient to fully recover out-of-plane motions. The downstream policy improves success rate by +13.5% compared to +6.2% for single-view LAM. A single cross-embodiment is sufficient for skill transfer, and adding a second viewpoint outperforms adding up to four further embodiments. The results transfer to hardware where demonstrations come from a human hand improving success rate by +20% over action-only policies in 850 rollouts. We show that video and robot data are complementary, and that benefits of learning from video are task-dependent. Website: this https URL
Abstract (translated)
URL
https://arxiv.org/abs/2610.27794
PDF
