# Research Papers — 2026-03 · bundle page 37 of 243 > Full-text markdown of papers #1801–1850 of 12123 published in 2026-03, newest first. AI-readable mirrors of the full 12123-paper 2026-03 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-03/page/37 [← Previous bundle](https://aipod.dev/papers/archive/2026-03/page/36.md) [Next bundle →](https://aipod.dev/papers/archive/2026-03/page/38.md) ### IncreRTL: Traceability-Guided Incremental RTL Generation under Requirement Evolution Authors: Luanrong Chen, Renzhi Chen, Xinyu Li, Shanshan Li, Rui Gong, Lei Wang Published: 2026-03-26 Categories: cs.SE, cs.AI, cs.AR arXiv: https://arxiv.org/abs/2603.25769v1 Large language models (LLMs) have shown promise in generating RTL code from natural-language descriptions, but existing methods remain static and struggle to adapt to evolving design requirements, potentially causing structural drift and costly full regeneration. We propose IncreRTL, a LLM-driven framework for incremental RTL generation under requirement evolution. By constructing requirement-code traceability links to locate and regenerate affected code segments, IncreRTL achieves accurate and consistent updates. Evaluated on our newly constructed EvoRTL-Bench, IncreRTL demonstrates notable improvements in regeneration consistency and efficiency, advancing LLM-based RTL generation toward practical engineering deployment. --- ### UCAgent: An End-to-End Agent for Block-Level Functional Verification Authors: Junyue Wang, Zhicheng Yao, Yan Pi, Xiaolong Li, Fangyuan Song, Jinru Wang, Yunlong Xie, Sa Wang, Yungang Bao Published: 2026-03-26 Categories: cs.SE, cs.AI, cs.AR, cs.MA arXiv: https://arxiv.org/abs/2603.25768v1 Functional verification remains a critical bottleneck in modern IC development cycles, accounting for approximately 70% of total development time in many projects. However, traditional methods, including constrained-random and formal verification, struggle to keep pace with the growing complexity of modern semiconductor designs. While recent advances in Large Language Models (LLMs) have shown promise in code generation and task automation, significant challenges hinder the realization of end-to-end functional verification automation. These challenges include (i) limited accuracy in generating Verilog/SystemVerilog verification code, (ii) the fragility of LLMs when executing complex, multi-step verification workflows, and (iii) the difficulty of maintaining verification consistency across specifications, coverage models, and test cases throughout the workflow. To address these challenges, we propose UCAgent, an end-to-end agent that automates hardware block-level functional verification based on three core mechanisms. First, we establish a pure Python verification environment using Picker and Toffee to avoid relying on LLM-generated SystemVerilog verification code. Second, we introduce a configurable 31-stage fine-grained verification workflow to guide the LLM, where each stage is verified by an automated checker. Furthermore, we propose a Verification Consistency Labeling Mechanism (VCLM) that assigns hierarchical labels to LLM-generated artifacts, improving the reliability and traceability of verification. Experimental results show that UCAgent can complete end-to-end automated verification on multiple modules, including the UART, FPU, and integer divider modules, achieving up to 98.5% code coverage and up to 100% functional coverage. UCAgent also discovers previously unidentified design defects in realistic designs, demonstrating its practical potential. --- ### Unlocking Strong Supervision: A Data-Centric Study of General-Purpose Audio Pre-Training Methods Authors: Xuanru Zhou, Yiwen Shao, Wei-Cheng Tseng, Dong Yu Published: 2026-03-26 Categories: cs.SD, cs.AI, eess.AS arXiv: https://arxiv.org/abs/2603.25767v1 Current audio pre-training seeks to learn unified representations for broad audio understanding tasks, but it remains fragmented and is fundamentally bottlenecked by its reliance on weak, noisy, and scale-limited labels. Drawing lessons from vision's foundational pre-training blueprint, we argue that the audio field must first establish its own large-scale, strong supervision framework. We introduce a new data-centric pipeline that leverages a high-fidelity captioner to create SOTA-quality captions and the first Unified Tag System (UTS) that bridges speech, music, and environmental sounds. We then conduct a systematic comparative study of different pre-training objectives on these strong source data. Our experiments suggest that data quality and coverage are the primary drivers of performance, while the choice of objective dictates downstream task specialization. --- ### ETA-VLA: Efficient Token Adaptation via Temporal Fusion and Intra-LLM Sparsification for Vision-Language-Action Models Authors: Yiru Wang, Anqing Jiang, Shuo Wang, Yuwen Heng, Zichong Gu, Hao Sun Published: 2026-03-26 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2603.25766v1 The integration of Vision-Language-Action (VLA) models into autonomous driving systems offers a unified framework for interpreting complex scenes and executing control commands. However, the necessity to incorporate historical multi-view frames for accurate temporal reasoning imposes a severe computational burden, primarily driven by the quadratic complexity of self-attention mechanisms in Large Language Models (LLMs). To alleviate this bottleneck, we propose ETA-VLA, an Efficient Token Adaptation framework for VLA models. ETA-VLA processes the past $n$ frames of multi-view images and introduces a novel Intra-LLM Sparse Aggregator (ILSA). Drawing inspiration from human driver attention allocation, ILSA dynamically identifies and prunes redundant visual tokens guided by textual queries and temporal consistency. Specifically, we utilize a text-guided scoring mechanism alongside a diversity-preserving sparsification strategy to select a sparse subset of critical tokens, ensuring comprehensive awareness of the driving scene. Extensive experiments on the NAVSIM v2 demonstrate that ETA-VLA achieves driving performance comparable to state-of-the-art baselines while reducing computational FLOPs by approximately 32\%. Notably, our method prunes 85% of visual tokens and reduces inference FLOPs by 61\%, but still retaining 94% of the original accuracy on the NAVSIM v2 benchmark. --- ### Evaluating Synthetic Images as Effective Substitutes for Experimental Data in Surface Roughness Classification Authors: Binwei Chen, Huachao Leng, Chi Yeung Mang, Tsz Wai Cheung, Yanhua Chen, Wai Keung Anthony Loh, Chi Ho Wong, Chak Yin Tang Published: 2026-03-26 Categories: cs.CV, cond-mat.mtrl-sci arXiv: https://arxiv.org/abs/2603.25765v1 Hard coatings play a critical role in industry, with ceramic materials offering outstanding hardness and thermal stability for applications that demand superior mechanical performance. However, deploying artificial intelligence (AI) for surface roughness classification is often constrained by the need for large labeled datasets and costly high-resolution imaging equipment. In this study, we explore the use of synthetic images, generated with Stable Diffusion XL, as an efficient alternative or supplement to experimentally acquired data for classifying ceramic surface roughness. We show that augmenting authentic datasets with generative images yields test accuracies comparable to those obtained using exclusively experimental images, demonstrating that synthetic images effectively reproduce the structural features necessary for classification. We further assess method robustness by systematically varying key training hyperparameters (epoch count, batch size, and learning rate), and identify configurations that preserve performance while reducing data requirements. Our results indicate that generative AI can substantially improve data efficiency and reliability in materials-image classification workflows, offering a practical route to lower experimental cost, accelerate model development, and expand AI applicability in materials engineering. --- ### Confident and Wrong: Silent Semantic Failures in Coding Agents Authors: Aman Mehta Published: 2026-03-26 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2603.25764v3 As coding agents move into production workflows, teams need to know not only whether an agent completes a task, but whether its action can be trusted. We show that completion and trustworthiness diverge sharply and systematically. Across 1,750 trajectories on 50 SWE-bench Verified tasks, we compare four frontier models over repeated runs and separate submit rate from test-verified resolve rate. GPT-5 submits a patch on 100% of runs but resolves only 44%; Llama 4 submits on 99% but resolves 18%; and Gemini, despite submitting least often at 70%, resolves more tasks than GPT-5 (50% versus 44%). These gaps are not random: they concentrate in one dangerous failure mode we call silent semantic failure. Qualitatively, on a buggy task the agent submits a plausible-looking patch on all five runs, yet none pass, the same misinterpretation repeated rather than random error. Quantitatively, it dominates failure, covering 80% of Llama 4's failing runs and 68% of GPT-5's, and it is invisible: the outcomes are confidently and consistently wrong, so completion-based and consistency-based monitoring both look healthy exactly when the agent should not be trusted. Lightweight pre-edit prompts do not close the gap. A second probe isolates the instinct to act: given an already-fixed bug, where the right move is to abstain, most models still edit the correct code. This action bias, acting when no action is warranted, is exactly what completion metrics reward. The throughline is measurement: submit rate captures action, but trust requires validity. So evaluation must catch up: score agents by test-verified correctness over repeated runs, report its uncertainty, and reward those that know when not to act. --- ### CANGuard: A Spatio-Temporal CNN-GRU-Attention Hybrid Architecture for Intrusion Detection in In-Vehicle CAN Networks Authors: Rakib Hossain Sajib, Md. Rokon Mia, Prodip Kumar Sarker, Abdullah Al Noman, Md Arifur Rahman Published: 2026-03-26 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2603.25763v1 The Internet of Vehicles (IoV) has become an essential component of smart transportation systems, enabling seamless interaction among vehicles and infrastructure. In recent years, it has played a progressively significant role in enhancing mobility, safety, and transportation efficiency. However, this connectivity introduces severe security vulnerabilities, particularly Denial-of-Service (DoS) and spoofing attacks targeting the Controller Area Network (CAN) bus, which could severely inhibit communication between the critical components of a vehicle, leading to system malfunctions, loss of control, or even endangering passengers' safety. To address this problem, this paper presents CANGuard, a novel spatio-temporal deep learning architecture that combines Convolutional Neural Networks (CNN), Gated Recurrent Units (GRU), and an attention mechanism to effectively identify such attacks. The model is trained and evaluated on the CICIoV2024 dataset, achieving competitive performance across accuracy, precision, recall, and F1-score and outperforming existing state-of-the-art methods. A comprehensive ablation study confirms the individual and combined contributions of the CNN, GRU, and attention components. Additionally, a SHAP analysis is conducted to interpret the decision-making process of the model and determine which features have the most significant impact on intrusion detection. The proposed approach demonstrates strong potential for practical and scalable security enhancements in modern IoV environments, thereby ensuring safer and more secure CAN bus communications. --- ### A Survey of OCR Evaluation Methods and Metrics and the Invisibility of Historical Documents Authors: Fitsum Sileshi Beyene, Christopher L. Dancy Published: 2026-03-26 Categories: cs.CV, cs.DL arXiv: https://arxiv.org/abs/2603.25761v1 Optical character recognition (OCR) and document understanding systems increasingly rely on large vision and vision-language models, yet evaluation remains centered on modern, Western, and institutional documents. This emphasis masks system behavior in historical and marginalized archives, where layout, typography, and material degradation shape interpretation. This study examines how OCR and document understanding systems are evaluated, with particular attention to Black historical newspapers. We review OCR and document understanding papers, as well as benchmark datasets, which are published between 2006 and 2025 using the PRISMA framework. We look into how the studies report training data, benchmark design, and evaluation metrics for vision transformer and multimodal OCR systems. During the review, we found that Black newspapers and other community-produced historical documents rarely appear in reported training data or evaluation benchmarks. Most evaluations emphasize character accuracy and task success on modern layouts. They rarely capture structural failures common in historical newspapers, including column collapse, typographic errors, and hallucinated text. To put these findings into perspective, we use previous empirical studies and archival statistics from significant Black press collections to show how evaluation gaps lead to structural invisibility and representational harm. We propose that these gaps occur due to organizational (meso) and institutional (macro) behaviors and structure, shaped by benchmark incentives and data governance decisions. --- ### ShotStream: Streaming Multi-Shot Video Generation for Interactive Storytelling Authors: Yawen Luo, Xiaoyu Shi, Junhao Zhuang, Yutian Chen, Quande Liu, Xintao Wang, Pengfei Wan, Tianfan Xue Published: 2026-03-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.25746v1 Multi-shot video generation is crucial for long narrative storytelling, yet current bidirectional architectures suffer from limited interactivity and high latency. We propose ShotStream, a novel causal multi-shot architecture that enables interactive storytelling and efficient on-the-fly frame generation. By reformulating the task as next-shot generation conditioned on historical context, ShotStream allows users to dynamically instruct ongoing narratives via streaming prompts. We achieve this by first fine-tuning a text-to-video model into a bidirectional next-shot generator, which is then distilled into a causal student via Distribution Matching Distillation. To overcome the challenges of inter-shot consistency and error accumulation inherent in autoregressive generation, we introduce two key innovations. First, a dual-cache memory mechanism preserves visual coherence: a global context cache retains conditional frames for inter-shot consistency, while a local context cache holds generated frames within the current shot for intra-shot consistency. And a RoPE discontinuity indicator is employed to explicitly distinguish the two caches to eliminate ambiguity. Second, to mitigate error accumulation, we propose a two-stage distillation strategy. This begins with intra-shot self-forcing conditioned on ground-truth historical shots and progressively extends to inter-shot self-forcing using self-generated histories, effectively bridging the train-test gap. Extensive experiments demonstrate that ShotStream generates coherent multi-shot videos with sub-second latency, achieving 16 FPS on a single GPU. It matches or exceeds the quality of slower bidirectional models, paving the way for real-time interactive storytelling. Training and inference code, as well as the models, are available on our --- ### Less Gaussians, Texture More: 4K Feed-Forward Textured Splatting Authors: Yixing Lao, Xuyang Bai, Xiaoyang Wu, Nuoyuan Yan, Zixin Luo, Tian Fang, Jean-Daniel Nahmias, Yanghai Tsin, Shiwei Li, Hengshuang Zhao Published: 2026-03-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.25745v1 Existing feed-forward 3D Gaussian Splatting methods predict pixel-aligned primitives, leading to a quadratic growth in primitive count as resolution increases. This fundamentally limits their scalability, making high-resolution synthesis such as 4K intractable. We introduce LGTM (Less Gaussians, Texture More), a feed-forward framework that overcomes this resolution scaling barrier. By predicting compact Gaussian primitives coupled with per-primitive textures, LGTM decouples geometric complexity from rendering resolution. This approach enables high-fidelity 4K novel view synthesis without per-scene optimization, a capability previously out of reach for feed-forward methods, all while using significantly fewer Gaussian primitives. Project page: https://yxlao.github.io/lgtm/ --- ### MuRF: Unlocking the Multi-Scale Potential of Vision Foundation Models Authors: Bocheng Zou, Mu Cai, Mark Stanley, Dingfu Lu, Yong Jae Lee Published: 2026-03-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.25744v2 Vision Foundation Models (VFMs) have become the cornerstone of modern computer vision, offering robust representations across a wide array of tasks. While recent advances allow these models to handle varying input sizes during training, inference typically remains restricted to a single, fixed scale. This prevalent single-scale paradigm overlooks a fundamental property of visual perception: varying resolutions offer complementary inductive biases, where low-resolution views excel at global semantic recognition and high-resolution views are essential for fine-grained refinement. In this work, we propose Multi-Resolution Fusion (MuRF), a simple yet universally effective strategy to harness this synergy at inference time. Instead of relying on a single view, MuRF constructs a unified representation by processing an image at multiple resolutions through a frozen VFM and fusing the resulting features. The universality of MuRF is its most compelling attribute. It is not tied to a specific architecture, serving instead as a fundamental, training-free enhancement to visual representation. We empirically validate this by applying MuRF to a broad spectrum of critical computer vision tasks across multiple distinct VFM families - primarily DINOv2, but also demonstrating successful generalization to contrastive models like SigLIP2. --- ### RefAlign: Representation Alignment for Reference-to-Video Generation Authors: Lei Wang, YuXin Song, Ge Wu, Haocheng Feng, Hang Zhou, Jingdong Wang, Yaxing Wang, Jian Yang Published: 2026-03-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.25743v2 Reference-to-video (R2V) generation is a controllable video synthesis paradigm that constrains the generation process using both text prompts and reference images, enabling applications such as personalized advertising and virtual try-on. In practice, existing R2V methods typically introduce additional high-level semantic or cross-modal features alongside the VAE latent representation of the reference image and jointly feed them into the diffusion Transformer (DiT). These auxiliary representations provide semantic guidance and act as implicit alignment signals, which can partially alleviate pixel-level information leakage in the VAE latent space. However, they may still struggle to address copy--paste artifacts and multi-subject confusion caused by modality mismatch across heterogeneous encoder features. In this paper, we propose RefAlign, a representation alignment framework that explicitly aligns DiT reference-branch features to the semantic space of a visual foundation model (VFM). The core of RefAlign is a reference alignment loss that pulls the reference features and VFM features of the same subject closer to improve identity consistency, while pushing apart the corresponding features of different subjects to enhance semantic discriminability. This simple yet effective strategy is applied only during training, incurring no inference-time overhead, and achieves a better balance between text controllability and reference fidelity. Extensive experiments on the OpenS2V-Eval benchmark demonstrate that RefAlign outperforms current state-of-the-art methods in TotalScore, validating the effectiveness of explicit reference alignment for R2V tasks. --- ### Vega: Learning to Drive with Natural Language Instructions Authors: Sicheng Zuo, Yuxuan Li, Wenzhao Zheng, Zheng Zhu, Jie Zhou, Jiwen Lu Published: 2026-03-26 Categories: cs.CV, cs.AI, cs.RO arXiv: https://arxiv.org/abs/2603.25741v2 Vision-language-action models have reshaped autonomous driving to incorporate languages into the decision-making process. However, most existing pipelines only utilize the language modality for scene descriptions or reasoning and lack the flexibility to follow diverse user instructions for personalized driving. To address this, we first construct a large-scale driving dataset (InstructScene) containing around 100,000 scenes annotated with diverse driving instructions with the corresponding trajectories. We then propose a unified Vision-Language-World-Action model, Vega, for instruction-based generation and planning. We employ the autoregressive paradigm to process visual inputs (vision) and language instructions (language) and the diffusion paradigm to generate future predictions (world modeling) and trajectories (action). We perform joint attention to enable interactions between the modalities and use individual projection layers for different modalities for more capabilities. Extensive experiments demonstrate that our method not only achieves superior planning performance but also exhibits strong instruction-following abilities, paving the way for more intelligent and personalized driving systems. --- ### Drive My Way: Preference Alignment of Vision-Language-Action Model for Personalized Driving Authors: Zehao Wang, Huaide Jiang, Shuaiwu Dong, Yuping Wang, Hang Qiu, Jiachen Li Published: 2026-03-26 Categories: cs.RO, cs.AI, cs.CV, cs.LG, cs.MA arXiv: https://arxiv.org/abs/2603.25740v1 Human driving behavior is inherently personal, which is shaped by long-term habits and influenced by short-term intentions. Individuals differ in how they accelerate, brake, merge, yield, and overtake across diverse situations. However, existing end-to-end autonomous driving systems either optimize for generic objectives or rely on fixed driving modes, lacking the ability to adapt to individual preferences or interpret natural language intent. To address this gap, we propose Drive My Way (DMW), a personalized Vision-Language-Action (VLA) driving framework that aligns with users' long-term driving habits and adapts to real-time user instructions. DMW learns a user embedding from our personalized driving dataset collected across multiple real drivers and conditions the policy on this embedding during planning, while natural language instructions provide additional short-term guidance. Closed-loop evaluation on the Bench2Drive benchmark demonstrates that DMW improves style instruction adaptation, and user studies show that its generated behaviors are recognizable as each driver's own style, highlighting personalization as a key capability for human-centered autonomous driving. Our data and code are available at https://dmw-cvpr.github.io/. --- ### MegaFlow: Zero-Shot Large Displacement Optical Flow Authors: Dingxi Zhang, Fangjinhua Wang, Marc Pollefeys, Haofei Xu Published: 2026-03-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.25739v2 Accurate estimation of large displacement optical flow remains a critical challenge. Existing methods typically rely on iterative local search or/and domain-specific fine-tuning, which severely limits their performance in large displacement and zero-shot generalization scenarios. To overcome this, we introduce MegaFlow, a simple yet powerful model for zero-shot large displacement optical flow. Rather than relying on highly complex, task-specific architectural designs, MegaFlow adapts powerful pre-trained vision priors to produce temporally consistent motion fields. In particular, we formulate flow estimation as a global matching problem by leveraging pre-trained global Vision Transformer features, which naturally capture large displacements. This is followed by a few lightweight iterative refinements to further improve the sub-pixel accuracy. Extensive experiments demonstrate that MegaFlow achieves state-of-the-art zero-shot performance across multiple optical flow benchmarks. Moreover, our model also delivers highly competitive zero-shot performance on long-range point tracking benchmarks, demonstrating its robust transferability and suggesting a unified paradigm for generalizable motion estimation. Our project page is at: https://kristen-z.github.io/projects/megaflow. --- ### PSDesigner: Automated Graphic Design with a Human-Like Creative Workflow Authors: Xincheng Shuai, Song Tang, Yutong Huang, Henghui Ding, Dacheng Tao Published: 2026-03-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.25738v1 Graphic design is a creative and innovative process that plays a crucial role in applications such as e-commerce and advertising. However, developing an automated design system that can faithfully translate user intentions into editable design files remains an open challenge. Although recent studies have leveraged powerful text-to-image models and MLLMs to assist graphic design, they typically simplify professional workflows, resulting in limited flexibility and intuitiveness. To address these limitations, we propose PSDesigner, an automated graphic design system that emulates the creative workflow of human designers. Building upon multiple specialized components, PSDesigner collects theme-related assets based on user instructions, and autonomously infers and executes tool calls to manipulate design files, such as integrating new assets or refining inferior elements. To endow the system with strong tool-use capabilities, we construct a design dataset, CreativePSD, which contains a large amount of high-quality PSD design files annotated with operation traces across a wide range of design scenarios and artistic styles, enabling models to learn expert design procedures. Extensive experiments demonstrate that PSDesigner outperforms existing methods across diverse graphic design tasks, empowering non-specialists to conveniently create production-quality designs. --- ### Training the Knowledge Base through Evidence Distillation and Write-Back Enrichment Authors: Yuxing Lu, Xukai Zhao, Wei Wu, Jinzhuo Wang Published: 2026-03-26 Categories: cs.AI, cs.CL, cs.IR arXiv: https://arxiv.org/abs/2603.25737v1 The knowledge base in a retrieval-augmented generation (RAG) system is typically assembled once and never revised, even though the facts a query requires are often fragmented across documents and buried in irrelevant content. We argue that the knowledge base should be treated as a trainable component and propose WriteBack-RAG, a framework that uses labeled examples to identify where retrieval succeeds, isolate the relevant documents, and distill them into compact knowledge units that are indexed alongside the original corpus. Because the method modifies only the corpus, it can be applied once as an offline preprocessing step and combined with any RAG pipeline. Across four RAG methods, six benchmarks, and two LLM backbones, WriteBack-RAG improves every evaluated setting, with gains averaging +2.14%. Cross-method transfer experiments further show that the distilled knowledge benefits RAG pipelines other than the one used to produce it, confirming that the improvement resides in the corpus itself. --- ### How good was my shot? Quantifying Player Skill Level in Table Tennis Authors: Akihiro Kubota, Tomoya Hasegawa, Ryo Kawahara, Ko Nishino Published: 2026-03-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.25736v1 Gauging an individual's skill level is crucial, as it inherently shapes their behavior. Quantifying skill, however, is challenging because it is latent to the observed actions. To explore skill understanding in human behavior, we focus on dyadic sports -- specifically table tennis -- where skill manifests not just in complex movements, but in the subtle nuances of execution conditioned on game context. Our key idea is to learn a generative model of each player's tactical racket strokes and jointly embed them in a common latent space that encodes individual characteristics, including those pertaining to skill levels. By training these player models on a large-scale dataset of 3D-reconstructed professional matches and conditioning them on comprehensive game context -- including player positioning and opponent behaviors -- the models capture individual tactical identities within their latent space. We probe this learned player space and find that it reflects distinct play styles and attributes that collectively represent skill. By training a simple relative ranking network on these embeddings, we demonstrate that both relative and absolute skill predictions can be achieved. These results demonstrate that the learned player space effectively quantifies skill levels, providing a foundation for automated skill assessment in complex, interactive behaviors. --- ### Unleashing Guidance Without Classifiers for Human-Object Interaction Animation Authors: Ziyin Wang, Sirui Xu, Chuan Guo, Bing Zhou, Jiangshan Gong, Jian Wang, Yu-Xiong Wang, Liang-Yan Gui Published: 2026-03-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.25734v1 Generating realistic human-object interaction (HOI) animations remains challenging because it requires jointly modeling dynamic human actions and diverse object geometries. Prior diffusion-based approaches often rely on hand-crafted contact priors or human-imposed kinematic constraints to improve contact quality. We propose LIGHT, a data-driven alternative in which guidance emerges from the denoising pace itself, reducing dependence on manually designed priors. Building on diffusion forcing, we factor the representation into modality-specific components and assign individualized noise levels with asynchronous denoising schedules. In this paradigm, cleaner components guide noisier ones through cross-attention, yielding guidance without auxiliary classifiers. We find that this data-driven guidance is inherently contact-aware, and can be enhanced when training is augmented with a broad spectrum of synthetic object geometries, encouraging invariance of contact semantics to geometric diversity. Extensive experiments show that pace-induced guidance more effectively mirrors the benefits of contact priors than conventional classifier-free guidance, while achieving higher contact fidelity, more realistic HOI generation, and stronger generalization to unseen objects and tasks. --- ### SlotVTG: Object-Centric Adapter for Generalizable Video Temporal Grounding Authors: Jiwook Han, Geo Ahn, Youngrae Kim, Jinwoo Choi Published: 2026-03-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.25733v1 Multimodal Large Language Models (MLLMs) have shown strong performance on Video Temporal Grounding (VTG). However, their coarse recognition capabilities are insufficient for fine-grained temporal understanding, making task-specific fine-tuning indispensable. This fine-tuning causes models to memorize dataset-specific shortcuts rather than faithfully grounding in the actual visual content, leading to poor Out-of-Domain (OOD) generalization. Object-centric learning offers a promising remedy by decomposing scenes into entity-level representations, but existing approaches require re-running the entire multi-stage training pipeline from scratch. We propose SlotVTG, a framework that steers MLLMs toward object-centric, input-grounded visual reasoning at minimal cost. SlotVTG introduces a lightweight slot adapter that decomposes visual tokens into abstract slots via slot attention and reconstructs the original sequence, where objectness priors from a self-supervised vision model encourage semantically coherent slot formation. Cross-domain evaluation on standard VTG benchmarks demonstrates that our approach significantly improves OOD robustness while maintaining competitive In-Domain (ID) performance with minimal overhead. --- ### BizGenEval: A Systematic Benchmark for Commercial Visual Content Generation Authors: Yan Li, Zezi Zeng, Ziwei Zhou, Xin Gao, Muzhao Tian, Yifan Yang, Mingxi Cheng, Qi Dai, Yuqing Yang, Lili Qiu, Zhendong Wang, Zhengyuan Yang et al. Published: 2026-03-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.25732v1 Recent advances in image generation models have expanded their applications beyond aesthetic imagery toward practical visual content creation. However, existing benchmarks mainly focus on natural image synthesis and fail to systematically evaluate models under the structured and multi-constraint requirements of real-world commercial design tasks. In this work, we introduce BizGenEval, a systematic benchmark for commercial visual content generation. The benchmark spans five representative document types: slides, charts, webpages, posters, and scientific figures, and evaluates four key capability dimensions: text rendering, layout control, attribute binding, and knowledge-based reasoning, forming 20 diverse evaluation tasks. BizGenEval contains 400 carefully curated prompts and 8000 human-verified checklist questions to rigorously assess whether generated images satisfy complex visual and semantic constraints. We conduct large-scale benchmarking on 26 popular image generation systems, including state-of-the-art commercial APIs and leading open-source models. The results reveal substantial capability gaps between current generative models and the requirements of professional visual content creation. We hope BizGenEval serves as a standardized benchmark for real-world commercial visual content generation. --- ### PackForcing: Short Video Training Suffices for Long Video Sampling and Long Context Inference Authors: Xiaofeng Mao, Shaohao Rui, Kaining Ying, Bo Zheng, Chuanhao Li, Mingmin Chi, Kaipeng Zhang Published: 2026-03-26 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2603.25730v1 Autoregressive video diffusion models have demonstrated remarkable progress, yet they remain bottlenecked by intractable linear KV-cache growth, temporal repetition, and compounding errors during long-video generation. To address these challenges, we present PackForcing, a unified framework that efficiently manages the generation history through a novel three-partition KV-cache strategy. Specifically, we categorize the historical context into three distinct types: (1) Sink tokens, which preserve early anchor frames at full resolution to maintain global semantics; (2) Mid tokens, which achieve a massive spatiotemporal compression (32x token reduction) via a dual-branch network fusing progressive 3D convolutions with low-resolution VAE re-encoding; and (3) Recent tokens, kept at full resolution to ensure local temporal coherence. To strictly bound the memory footprint without sacrificing quality, we introduce a dynamic top-$k$ context selection mechanism for the mid tokens, coupled with a continuous Temporal RoPE Adjustment that seamlessly re-aligns position gaps caused by dropped tokens with negligible overhead. Empowered by this principled hierarchical context compression, PackForcing can generate coherent 2-minute, 832x480 videos at 16 FPS on a single H200 GPU. It achieves a bounded KV cache of just 4 GB and enables a remarkable 24x temporal extrapolation (5s to 120s), operating effectively either zero-shot or trained on merely 5-second clips. Extensive results on VBench demonstrate state-of-the-art temporal consistency (26.07) and dynamic degree (56.25), proving that short-video supervision is sufficient for high-quality, long-video synthesis. https://github.com/ShandaAI/PackForcing --- ### PixelSmile: Toward Fine-Grained Facial Expression Editing Authors: Jiabin Hua, Hengyuan Xu, Aojie Li, Wei Cheng, Gang Yu, Xingjun Ma, Yu-Gang Jiang Published: 2026-03-26 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2603.25728v1 Fine-grained facial expression editing has long been limited by intrinsic semantic overlap. To address this, we construct the Flex Facial Expression (FFE) dataset with continuous affective annotations and establish FFE-Bench to evaluate structural confusion, editing accuracy, linear controllability, and the trade-off between expression editing and identity preservation. We propose PixelSmile, a diffusion framework that disentangles expression semantics via fully symmetric joint training. PixelSmile combines intensity supervision with contrastive learning to produce stronger and more distinguishable expressions, achieving precise and stable linear expression control through textual latent interpolation. Extensive experiments demonstrate that PixelSmile achieves superior disentanglement and robust identity preservation, confirming its effectiveness for continuous, controllable, and fine-grained expression editing, while naturally supporting smooth expression blending. --- ### Back to Basics: Revisiting ASR in the Age of Voice Agents Authors: Geeyang Tay, Wentao Ma, Jaewon Lee, Yuzhi Tang, Daniel Lee, Weisu Yin, Dongming Shen, Silin Meng, Yi Zhu, Mu Li, Alex Smola Published: 2026-03-26 Categories: cs.AI, cs.MM arXiv: https://arxiv.org/abs/2603.25727v1 Automatic speech recognition (ASR) systems have achieved near-human accuracy on curated benchmarks, yet still fail in real-world voice agents under conditions that current evaluations do not systematically cover. Without diagnostic tools that isolate specific failure factors, practitioners cannot anticipate which conditions, in which languages, will cause what degree of degradation. We introduce WildASR, a multilingual (four-language) diagnostic benchmark sourced entirely from real human speech that factorizes ASR robustness along three axes: environmental degradation, demographic shift, and linguistic diversity. Evaluating seven widely used ASR systems, we find severe and uneven performance degradation, and model robustness does not transfer across languages or conditions. Critically, models often hallucinate plausible but unspoken content under partial or degraded inputs, creating concrete safety risks for downstream agent behavior. Our results demonstrate that targeted, factor-isolated evaluation is essential for understanding and improving ASR reliability in production systems. Besides the benchmark itself, we also present three analytical tools that practitioners can use to guide deployment decisions. --- ### AnyHand: A Large-Scale Synthetic Dataset for RGB(-D) Hand Pose Estimation Authors: Chen Si, Yulin Liu, Bo Ai, Jianwen Xie, Rolandos Alexandros Potamias, Chuanxia Zheng, Hao Su Published: 2026-03-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.25726v3 We present AnyHand, a large-scale synthetic dataset designed to advance the state of the art in 3D hand pose estimation. While recent works with foundation approaches have shown that scaling training data markedly improves hand pose estimation, existing real-world datasets are limited in coverage, and prior synthetic datasets rarely provide occlusions, arm details, and aligned depth together at scale. To address this bottleneck, our proposed AnyHand contains 2.5M single-hand and 4.1M hand-object interaction RGB-D images, with rich geometric annotations. We show that extending the original training data recipes of existing RGB baselines with AnyHand yields significant gains on multiple benchmarks (FreiHAND and HO-3D), even when keeping the architectures and training schemes fixed. Together with extensive ablations on the scale and composition of the training data setups, these results suggest that training data diversity and quality are as critical as scale for advancing hand pose estimation. We further examine the utility of AnyHand's aligned depth maps in the appendix, showing that scaling RGB-D supervision with AnyHand allows a lightweight depth-fusion variant of existing RGB baselines to outperform prior RGB-D methods. --- ### Natural-Language Agent Harnesses Authors: Linyue Pan, Lexiao Zou, Shuo Guo, Jingchen Ni, Hai-Tao Zheng Published: 2026-03-26 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2603.25723v2 Agent performance is strongly shaped by the surrounding harness: the external execution system around a model that organizes a task run. Yet this logic is usually buried in tightly coupled controller code, which makes harnesses hard to inspect, compare, transfer, and ablate. This paper asks whether the reusable design pattern of an agent harness can be represented as an executable natural-language object. We introduce Natural-Language Agent Harnesses (NLAHs), editable documents that describe run-level harness policy, and Intelligent Harness Runtime (IHR), a shared runtime that interprets these documents into agent calls, handoffs, state updates, validation gates, and artifact contracts. Across coding, terminal-use, and computer-use benchmarks, IHR-executed NLAHs achieve comparable task outcomes to code and prompted realizations, while exposing much shorter static harness policies. Module ablations further show that explicit harness modules are analyzable. These results suggest that agent harnesses can be turned from incidental glue around models into scientific representation objects. --- ### No Hard Negatives Required: Concept Centric Learning Leads to Compositionality without Degrading Zero-shot Capabilities of Contrastive Models Authors: Hai X. Pham, David T. Hoffmann, Ricardo Guerrero, Brais Martinez Published: 2026-03-26 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2603.25722v2 Contrastive vision-language (V&L) models remain a popular choice for various applications. However, several limitations have emerged, most notably the limited ability of V&L models to learn compositional representations. Prior methods often addressed this limitation by generating custom training data to obtain hard negative samples. Hard negatives have been shown to improve performance on compositionality tasks, but are often specific to a single benchmark, do not generalize, and can cause substantial degradation of basic V&L capabilities such as zero-shot or retrieval performance, rendering them impractical. In this work we follow a different approach. We identify two root causes that limit compositionality performance of V&Ls: 1) Long training captions do not require a compositional representation; and 2) The final global pooling in the text and image encoders lead to a complete loss of the necessary information to learn binding in the first place. As a remedy, we propose two simple solutions: 1) We obtain short concept centric caption parts using standard NLP software and align those with the image; and 2) We introduce a parameter-free cross-modal attention-pooling to obtain concept centric visual embeddings from the image encoder. With these two changes and simple auxiliary contrastive losses, we obtain SOTA performance on standard compositionality benchmarks, while maintaining or improving strong zero-shot and retrieval capabilities. This is achieved without increasing inference cost. We release the code for this work at https://github.com/saic-fi/concept_centric_clip. --- ### R-C2: Cycle-Consistent Reinforcement Learning Improves Multimodal Reasoning Authors: Zirui Zhang, Haoyu Dong, Kexin Pei, Chengzhi Mao Published: 2026-03-26 Categories: cs.AI, cs.CV arXiv: https://arxiv.org/abs/2603.25720v1 Robust perception and reasoning require consistency across sensory modalities. Yet current multimodal models often violate this principle, yielding contradictory predictions for visual and textual representations of the same concept. Rather than masking these failures with standard voting mechanisms, which can amplify systematic biases, we show that cross-modal inconsistency provides a rich and natural signal for learning. We introduce RC2, a reinforcement learning framework that resolves internal conflicts by enforcing cross-modal cycle consistency. By requiring a model to perform backward inference, switch modalities, and reliably reconstruct the answer through forward inference, we obtain a dense, label-free reward. This cyclic constraint encourages the model to align its internal representations autonomously. Optimizing for this structure mitigates modality-specific errors and improves reasoning accuracy by up to 7.6 points. Our results suggest that advanced reasoning emerges not only from scaling data, but also from enforcing a structurally consistent understanding of the world. --- ### Agent Factories for High Level Synthesis: How Far Can General-Purpose Coding Agents Go in Hardware Optimization? Authors: Abhishek Bhandwaldar, Mihir Choudhury, Ruchir Puri, Akash Srivastava Published: 2026-03-26 Categories: cs.AI, cs.AR, cs.LG arXiv: https://arxiv.org/abs/2603.25719v2 We present an empirical study of how far general-purpose coding agents -- without hardware-specific training -- can optimize hardware designs from high-level algorithmic specifications. We introduce an agent factory, a two-stage pipeline that constructs and coordinates multiple autonomous optimization agents. In Stage~1, the pipeline decomposes a design into sub-kernels, independently optimizes each using pragma and code-level transformations, and formulates an Integer Linear Program (ILP) to assemble globally promising configurations under an area constraint. In Stage~2, it launches $N$ expert agents over the top ILP solutions, each exploring cross-function optimizations such as pragma recombination, loop fusion, and memory restructuring that are not captured by sub-kernel decomposition. We evaluate the approach on 12 kernels from HLS-Eval and Rodinia-HLS using Claude Code (Opus~4.5/4.6) with AMD Vitis HLS. Scaling from 1 to 10 agents yields a mean $8.27\times$ speedup over baseline, with larger gains on harder benchmarks: streamcluster exceeds $20\times$ and kmeans reaches approximately $10\times$. Across benchmarks, agents consistently rediscover known hardware optimization patterns without domain-specific training, and the best designs often do not originate from top-ranked ILP candidates, indicating that global optimization exposes improvements missed by sub-kernel search. These results establish agent scaling as a practical and effective axis for HLS optimization. --- ### Out of Sight but Not Out of Mind: Hybrid Memory for Dynamic Video World Models Authors: Kaijin Chen, Dingkang Liang, Xin Zhou, Yikang Ding, Xiaoqiang Liu, Pengfei Wan, Xiang Bai Published: 2026-03-26 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2603.25716v2 Video world models have shown immense potential in simulating the physical world, yet existing memory mechanisms primarily treat environments as static canvases. When dynamic subjects hide out of sight and later re-emerge, current methods often struggle, leading to frozen, distorted, or vanishing subjects. To address this, we introduce Hybrid Memory, a novel paradigm requiring models to simultaneously act as precise archivists for static backgrounds and vigilant trackers for dynamic subjects, ensuring motion continuity during out-of-view intervals. To facilitate research in this direction, we construct HM-World, the first large-scale video dataset dedicated to hybrid memory. It features 59K high-fidelity clips with decoupled camera and subject trajectories, encompassing 17 diverse scenes, 49 distinct subjects, and meticulously designed exit-entry events to rigorously evaluate hybrid coherence. Furthermore, we propose HyDRA, a specialized memory architecture that compresses memory into tokens and utilizes a spatiotemporal relevance-driven retrieval mechanism. By selectively attending to relevant motion cues, HyDRA effectively preserves the identity and motion of hidden subjects. Extensive experiments on HM-World demonstrate that our method significantly outperforms state-of-the-art approaches in both dynamic subject consistency and overall generation quality. Code is publicly available at https://github.com/H-EmbodVis/HyDRA. --- ### Seeing to Ground: Visual Attention for Hallucination-Resilient MDLLMs Authors: Vishal Narnaware, Animesh Gupta, Kevin Zhai, Zhenyi Wang, Mubarak Shah Published: 2026-03-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.25711v1 Multimodal Diffusion Large Language Models (MDLLMs) achieve high-concurrency generation through parallel masked decoding, yet the architectures remain prone to multimodal hallucinations. This structural vulnerability stems from an algorithmic flaw: the decoder ranks candidate tokens based on textual likelihood without verifying localized visual support. We establish that this language-only ranking induces an objective mismatch, where language probability mass acts as a misspecified proxy for the intended multimodal task. Consequently, we reinterpret hallucination as a localized optimization error, a phenomenon where the decoder exploits language shortcuts to maximize a proxy score at the expense of visual grounding. To address this objective mismatch, we introduce VISAGE, a training-free decoding framework that calibrates the objective at inference time. VISAGE estimates the proxy discrepancy by quantifying the spatial entropy of cross-attention distributions. By enforcing a localization consensus across attention heads, the method penalizes spatially uniform distributions and re-ranks token commitments to favor visually grounded outcomes. We provide an analytical stability guarantee establishing that VISAGE maintains a bounded objective loss under estimation error. Evaluations across hallucination-sensitive and general-purpose benchmarks demonstrate the robustness of the framework, yielding relative gains of 8.59% on MMMU-val and 7.75% on HallusionBench. --- ### TRACE: Object Motion Editing in Videos with First-Frame Trajectory Guidance Authors: Quynh Phung, Long Mai, Cusuh Ham, Feng Liu, Jia-Bin Huang, Aniruddha Mahapatra Published: 2026-03-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.25707v1 We study object motion path editing in videos, where the goal is to alter a target object's trajectory while preserving the original scene content. Unlike prior video editing methods that primarily manipulate appearance or rely on point-track-based trajectory control, which is often challenging for users to provide during inference, especially in videos with camera motion, we offer a practical, easy-to-use approach to controllable object-centric motion editing. We present Trace, a framework that enables users to design the desired trajectory in a single anchor frame and then synthesizes a temporally consistent edited video. Our approach addresses this task with a two-stage pipeline: a cross-view motion transformation module that maps first-frame path design to frame-aligned box trajectories under camera motion, and a motion-conditioned video re-synthesis module that follows these trajectories to regenerate the object while preserving the remaining content of the input video. Experiments on diverse real-world videos show that our method produces more coherent, realistic, and controllable motion edits than recent image-to-video and video-to-video methods. --- ### Wan-Weaver: Interleaved Multi-modal Generation via Decoupled Training Authors: Jinbo Xing, Zeyinzi Jiang, Yuxiang Tuo, Chaojie Mao, Xiaotang Gai, Xi Chen, Jingfeng Zhang, Yulin Pan, Zhen Han, Jie Xiao, Keyu Yan, Chenwei Xie et al. Published: 2026-03-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.25706v2 Recent unified models have made unprecedented progress in both understanding and generation. However, while most of them accept multi-modal inputs, they typically produce only single-modality outputs. This challenge of producing interleaved content is mainly due to training data scarcity and the difficulty of modeling long-range cross-modal context. To address this issue, we decompose interleaved generation into textual planning and visual consistency modeling, and introduce a framework consisting of a planner and a visualizer. The planner produces dense textual descriptions for visual content, while the visualizer synthesizes images accordingly. Under this guidance, we construct large-scale textual-proxy interleaved data (where visual content is represented in text) to train the planner, and curate reference-guided image data to train the visualizer. These designs give rise to Wan-Weaver, which exhibits emergent interleaved generation ability with long-range textual coherence and visual consistency. Meanwhile, the integration of diverse understanding and generation data into planner training enables Wan-Weaver to achieve robust task reasoning and generation proficiency. To assess the model's capability in interleaved generation, we further construct a benchmark that spans a wide range of use cases across multiple dimensions. Extensive experiments demonstrate that, even without access to any real interleaved data, Wan-Weaver achieves superior performance over existing methods. --- ### S2D2: Fast Decoding for Diffusion LLMs via Training-Free Self-Speculation Authors: Ligong Han, Hao Wang, Han Gao, Kai Xu, Akash Srivastava Published: 2026-03-26 Categories: cs.CL arXiv: https://arxiv.org/abs/2603.25702v2 Block-diffusion language models offer a promising path toward faster-than-autoregressive generation by combining block-wise autoregressive decoding with within-block parallel denoising. However, in the few-step regime needed for practical acceleration, standard confidence-thresholded decoding is often brittle: aggressive thresholds hurt quality, while conservative thresholds require unnecessary denoising steps. Existing approaches that address this issue either require additional training or incur extra test-time compute. We present S2D2, a training-free self-speculative decoding framework for block-diffusion language models. Our key observation is that a block-diffusion model becomes autoregressive when the block size is reduced to one, allowing the same pretrained model to act as both drafter and verifier. S2D2 inserts a speculative verification step into standard block-diffusion decoding and uses lightweight routing policies to decide when verification is worth its cost. This yields a hybrid decoding trajectory in which diffusion proposes tokens in parallel, while the autoregressive mode acts as a local sequence-level critic. Across three mainstream block-diffusion families, S2D2 consistently improves the accuracy-speed tradeoff over strong confidence-thresholding baselines. On SDAR, we observe up to $4.7\times$ speedup over autoregressive decoding, and up to $1.57\times$ over a tuned dynamic decoding baseline while improving accuracy by up to $4.5$ points. On LLaDA2.1-Mini, S2D2 remains complementary to built-in self-correction, including a conservative setting where it is $4.4\times$ faster than the static baseline with slightly higher accuracy. --- ### Neural Network Conversion of Machine Learning Pipelines Authors: Man-Ling Sung, Jan Silovsky, Man-Hung Siu, Herbert Gish, Chinnu Pittapally Published: 2026-03-26 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2603.25699v1 Transfer learning and knowledge distillation has recently gained a lot of attention in the deep learning community. One transfer approach, the student-teacher learning, has been shown to successfully create ``small'' student neural networks that mimic the performance of a much bigger and more complex ``teacher'' networks. In this paper, we investigate an extension to this approach and transfer from a non-neural-based machine learning pipeline as teacher to a neural network (NN) student, which would allow for joint optimization of the various pipeline components and a single unified inference engine for multiple ML tasks. In particular, we explore replacing the random forest classifier by transfer learning to a student NN. We experimented with various NN topologies on 100 OpenML tasks in which random forest has been one of the best solutions. Our results show that for the majority of the tasks, the student NN can indeed mimic the teacher if one can select the right NN hyper-parameters. We also investigated the use of random forest for selecting the right NN hyper-parameters. --- ### The Kitchen Loop: User-Spec-Driven Development for a Self-Evolving Codebase Authors: Yannick Roy Published: 2026-03-26 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2603.25697v1 Code production is now a commodity; the bottleneck is knowing what to build and proving it works. We present the Kitchen Loop, a framework for autonomous, self-evolving software built on a unified trust model: (1) a specification surface enumerating what the product claims to support; (2) 'As a User x 1000', where an LLM agent exercises that surface as a synthetic power user at 1,000x human cadence; (3) Unbeatable Tests, ground-truth verification the code author cannot fake; and (4) Drift Control, continuous quality measurement with automated pause gates. We validate across two production systems over 285+ iterations, producing 1,094+ merged pull requests with zero regressions detected by the regression oracle (methodology in Section 6.1). We observe emergent properties at scale: multi-iteration self-correction chains, autonomous infrastructure healing, and monotonically improving quality gates. The primitives are not new; our contribution is their composition into a production-tested system with the operational discipline that makes long-running autonomous evolution safe. --- ### A Unified Memory Perspective for Probabilistic Trustworthy AI Authors: Xueji Zhao, Likai Pei, Jianbo Liu, Kai Ni, Ningyuan Cao Published: 2026-03-26 Categories: cs.LG, cs.AI, cs.AR, cs.ET arXiv: https://arxiv.org/abs/2603.25692v1 Trustworthy artificial intelligence increasingly relies on probabilistic computation to achieve robustness, interpretability, security and privacy. In practical systems, such workloads interleave deterministic data access with repeated stochastic sampling across models, data paths and system functions, shifting performance bottlenecks from arithmetic units to memory systems that must deliver both data and randomness. Here we present a unified data-access perspective in which deterministic access is treated as a limiting case of stochastic sampling, enabling both modes to be analyzed within a common framework. This view reveals that increasing stochastic demand reduces effective data-access efficiency and can drive systems into entropy-limited operation. Based on this insight, we define memory-level evaluation criteria, including unified operation, distribution programmability, efficiency, robustness to hardware non-idealities and parallel compatibility. Using these criteria, we analyze limitations of conventional architectures and examine emerging probabilistic compute-in-memory approaches that integrate sampling with memory access, outlining pathways toward scalable hardware for trustworthy AI. --- ### LEMMA: Laplacian pyramids for Efficient Marine SeMAntic Segmentation Authors: Ishaan Gakhar, Laven Srivastava, Sankarshanaa Sagaram, Aditya Kasliwal, Ujjwal Verma Published: 2026-03-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.25689v1 Semantic segmentation in marine environments is crucial for the autonomous navigation of unmanned surface vessels (USVs) and coastal Earth Observation events such as oil spills. However, existing methods, often relying on deep CNNs and transformer-based architectures, face challenges in deployment due to their high computational costs and resource-intensive nature. These limitations hinder the practicality of real-time, low-cost applications in real-world marine settings. To address this, we propose LEMMA, a lightweight semantic segmentation model designed specifically for accurate remote sensing segmentation under resource constraints. The proposed architecture leverages Laplacian Pyramids to enhance edge recognition, a critical component for effective feature extraction in complex marine environments for disaster response, environmental surveillance, and coastal monitoring. By integrating edge information early in the feature extraction process, LEMMA eliminates the need for computationally expensive feature map computations in deeper network layers, drastically reducing model size, complexity and inference time. LEMMA demonstrates state-of-the-art performance across datasets captured from diverse platforms while reducing trainable parameters and computational requirements by up to 71x, GFLOPs by up to 88.5\%, and inference time by up to 84.65\%, as compared to existing models. Experimental results highlight its effectiveness and real-world applicability, including 93.42\% IoU on the Oil Spill dataset and 98.97\% mIoU on Mastr1325. --- ### On Neural Scaling Laws for Weather Emulation through Continual Training Authors: Shashank Subramanian, Alexander Kiefer, Arnur Nigmetov, Amir Gholami, Dmitriy Morozov, Michael W. Mahoney Published: 2026-03-26 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.25687v1 Neural scaling laws, which in some domains can predict the performance of large neural networks as a function of model, data, and compute scale, are the cornerstone of building foundation models in Natural Language Processing and Computer Vision. We study neural scaling in Scientific Machine Learning, focusing on models for weather forecasting. To analyze scaling behavior in as simple a setting as possible, we adopt a minimal, scalable, general-purpose Swin Transformer architecture, and we use continual training with constant learning rates and periodic cooldowns as an efficient training strategy. We show that models trained in this minimalist way follow predictable scaling trends and even outperform standard cosine learning rate schedules. Cooldown phases can be re-purposed to improve downstream performance, e.g., enabling accurate multi-step rollouts over longer forecast horizons as well as sharper predictions through spectral loss adjustments. We also systematically explore a wide range of model and dataset sizes under various compute budgets to construct IsoFLOP curves, and we identify compute-optimal training regimes. Extrapolating these trends to larger scales highlights potential performance limits, demonstrating that neural scaling can serve as an important diagnostic for efficient resource allocation. We open-source our code for reproducibility. --- ### Just Zoom In: Cross-View Geo-Localization via Autoregressive Zooming Authors: Yunus Talha Erzurumlu, Jiyong Kwag, Alper Yilmaz Published: 2026-03-26 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2603.25686v1 Cross-view geo-localization (CVGL) estimates a camera's location by matching a street-view image to geo-referenced overhead imagery, enabling GPS-denied localization and navigation. Existing methods almost universally formulate CVGL as an image-retrieval problem in a contrastively trained embedding space. This ties performance to large batches and hard negative mining, and it ignores both the geometric structure of maps and the coverage mismatch between street-view and overhead imagery. In particular, salient landmarks visible from the street view can fall outside a fixed satellite crop, making retrieval targets ambiguous and limiting explicit spatial inference over the map. We propose Just Zoom In, an alternative formulation that performs CVGL via autoregressive zooming over a city-scale overhead map. Starting from a coarse satellite view, the model takes a short sequence of zoom-in decisions to select a terminal satellite cell at a target resolution, without contrastive losses or hard negative mining. We further introduce a realistic benchmark with crowd-sourced street views and high-resolution satellite imagery that reflects real capture conditions. On this benchmark, Just Zoom In achieves state-of-the-art performance, improving Recall@1 within 50 m by 5.5% and Recall@1 within 100 m by 9.6% over the strongest contrastive-retrieval baseline. These results demonstrate the effectiveness of sequential coarse-to-fine spatial reasoning for cross-view geo-localization. --- ### Persistent Robot World Models: Stabilizing Multi-Step Rollouts via Reinforcement Learning Authors: Jai Bardhan, Patrik Drozdik, Josef Sivic, Vladimir Petrik Published: 2026-03-26 Categories: cs.RO, cs.CV arXiv: https://arxiv.org/abs/2603.25685v2 Action-conditioned robot world models generate future video frames of the manipulated scene given a robot action sequence, offering a promising alternative for simulating tasks that are difficult to model with traditional physics engines. However, these models are optimized for short-term prediction and break down when deployed autoregressively: each predicted clip feeds back as context for the next, causing errors to compound and visual quality to rapidly degrade. We address this through the following contributions. First, we introduce a reinforcement learning (RL) post-training scheme that trains the world model on its own autoregressive rollouts rather than on ground-truth histories. We achieve this by adapting a recent contrastive RL objective for diffusion models to our setting and show that its convergence guarantees carry over exactly. Second, we design a training protocol that generates and compares multiple candidate variable-length futures from the same rollout state, reinforcing higher-fidelity predictions over lower-fidelity ones. Third, we develop efficient, multi-view visual fidelity rewards that combine complementary perceptual metrics across camera views and are aggregated at the clip level for dense, low-variance training signal. Fourth, we show that our approach establishes a new state-of-the-art for rollout fidelity on the DROID dataset, outperforming the strongest baseline on all metrics (e.g., LPIPS reduced by 14% on external cameras, SSIM improved by 9.1% on the wrist camera), winning 98% of paired comparisons, and achieving an 80% preference rate in a blind human study. --- ### Persistent Robot World Models: Stabilizing Multi-Step Rollouts via Reinforcement Learning Authors: Jai Bardhan, Patrik Drozdik, Josef Sivic, Vladimir Petrik Published: 2026-03-26 Categories: cs.RO, cs.CV arXiv: https://arxiv.org/abs/2603.25685v1 Action-conditioned robot world models generate future video frames of the manipulated scene given a robot action sequence, offering a promising alternative for simulating tasks that are difficult to model with traditional physics engines. However, these models are optimized for short-term prediction and break down when deployed autoregressively: each predicted clip feeds back as context for the next, causing errors to compound and visual quality to rapidly degrade. We address this through the following contributions. First, we introduce a reinforcement learning (RL) post-training scheme that trains the world model on its own autoregressive rollouts rather than on ground-truth histories. We achieve this by adapting a recent contrastive RL objective for diffusion models to our setting and show that its convergence guarantees carry over exactly. Second, we design a training protocol that generates and compares multiple candidate variable-length futures from the same rollout state, reinforcing higher-fidelity predictions over lower-fidelity ones. Third, we develop efficient, multi-view visual fidelity rewards that combine complementary perceptual metrics across camera views and are aggregated at the clip level for dense, low-variance training signal. Fourth, we show that our approach establishes a new state-of-the-art for rollout fidelity on the DROID dataset, outperforming the strongest baseline on all metrics (e.g., LPIPS reduced by 14% on external cameras, SSIM improved by 9.1% on the wrist camera), winning 98% of paired comparisons, and achieving an 80% preference rate in a blind human study. --- ### Self-Improvement of Large Language Models: A Technical Overview and Future Outlook Authors: Haoyan Yang, Mario Xerri, Solha Park, Huajian Zhang, Yiyang Feng, Sai Akhil Kogilathota, Jiawei Zhou Published: 2026-03-26 Categories: cs.CL arXiv: https://arxiv.org/abs/2603.25681v3 As large language models (LLMs) continue to advance, improving them solely through human supervision is becoming increasingly costly and limited in scalability. As models approach human-level capabilities in certain domains, human feedback may no longer provide sufficiently informative signals for further improvement. At the same time, the growing ability of models to make autonomous decisions and execute complex actions naturally enables abstractions in which components of the model development process can be progressively automated. Together, these challenges and opportunities have driven increasing interest in self-improvement, where models autonomously generate data, evaluate outputs, and iteratively refine their own capabilities. In this paper, we present a system-level perspective on self-improving language models and introduce a unified framework that organizes existing techniques. We conceptualize the self-improvement system as a closed-loop lifecycle, consisting of four tightly coupled processes: data acquisition, data selection, model optimization, and inference refinement, along with an autonomous evaluation layer throughout the process. Within this framework, the model itself plays a central role in driving each stage: collecting or generating data, selecting informative signals, updating its parameters, and refining outputs, while the autonomous evaluation layer continuously monitors progress and guides the improvement cycle across stages. Following this lifecycle perspective, we systematically review and analyze representative methods for each component from a technical standpoint. We further examine current limitations, potential risks, and prominent applications, and outline our vision for future research toward fully self-improving LLMs. --- ### Self-Improvement of Large Language Models: A Technical Overview and Future Outlook Authors: Haoyan Yang, Mario Xerri, Solha Park, Huajian Zhang, Yiyang Feng, Sai Akhil Kogilathota, Jiawei Zhou Published: 2026-03-26 Categories: cs.CL arXiv: https://arxiv.org/abs/2603.25681v1 As large language models (LLMs) continue to advance, improving them solely through human supervision is becoming increasingly costly and limited in scalability. As models approach human-level capabilities in certain domains, human feedback may no longer provide sufficiently informative signals for further improvement. At the same time, the growing ability of models to make autonomous decisions and execute complex actions naturally enables abstractions in which components of the model development process can be progressively automated. Together, these challenges and opportunities have driven increasing interest in self-improvement, where models autonomously generate data, evaluate outputs, and iteratively refine their own capabilities. In this paper, we present a system-level perspective on self-improving language models and introduce a unified framework that organizes existing techniques. We conceptualize the self-improvement system as a closed-loop lifecycle, consisting of four tightly coupled processes: data acquisition, data selection, model optimization, and inference refinement, along with an autonomous evaluation layer. Within this framework, the model itself plays a central role in driving each stage: collecting or generating data, selecting informative signals, updating its parameters, and refining outputs, while the autonomous evaluation layer continuously monitors progress and guides the improvement cycle across stages. Following this lifecycle perspective, we systematically review and analyze representative methods for each component from a technical standpoint. We further discuss current limitations and outline our vision for future research toward fully self-improving LLMs. --- ### Measuring What Matters -- or What's Convenient?: Robustness of LLM-Based Scoring Systems to Construct-Irrelevant Factors Authors: Cole Walsh, Rodica Ivan Published: 2026-03-26 Categories: cs.CL, cs.AI, cs.CY arXiv: https://arxiv.org/abs/2603.25674v1 Automated systems have been widely adopted across the educational testing industry for open-response assessment and essay scoring. These systems commonly achieve performance levels comparable to or superior than trained human raters, but have frequently been demonstrated to be vulnerable to the influence of construct-irrelevant factors (i.e., features of responses that are unrelated to the construct assessed) and adversarial conditions. Given the rising usage of large language models in automated scoring systems, there is a renewed focus on ``hallucinations'' and the robustness of these LLM-based automated scoring approaches to construct-irrelevant factors. This study investigates the effects of construct-irrelevant factors on a dual-architecture LLM-based scoring system designed to score short essay-like open-response items in a situational judgment test. It was found that the scoring system was generally robust to padding responses with meaningless text, spelling errors, and writing sophistication. Duplicating large passages of text resulted in lower scores predicted by the system, on average, contradicting results from previous studies of non-LLM-based scoring systems, while off-topic responses were heavily penalized by the scoring system. These results provide encouraging support for the robustness of future LLM-based scoring systems when designed with construct relevance in mind. --- ### Longitudinal Digital Phenotyping for Early Cognitive-Motor Screening Authors: Diego Jimenez-Oviedo, Ruben Vera-Rodriguez, Ruben Tolosana, Juan Carlos Ruiz-Garcia, Jaime Herreros-Rodriguez Published: 2026-03-26 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.25673v1 Early detection of atypical cognitive-motor development is critical for timely intervention, yet traditional assessments rely heavily on subjective, static evaluations. The integration of digital devices offers an opportunity for continuous, objective monitoring through digital biomarkers. In this work, we propose an AI-driven longitudinal framework to model developmental trajectories in children aged 18 months to 8 years. Using a dataset of tablet-based interactions collected over multiple academic years, we analyzed six cognitive-motor tasks (e.g., fine motor control, reaction time). We applied dimensionality reduction (t-SNE) and unsupervised clustering (K-Means++) to identify distinct developmental phenotypes and tracked individual transitions between these profiles over time. Our analysis reveals three distinct profiles: low, medium, and high performance. Crucially, longitudinal tracking highlights a high stability in the low-performance cluster (>90% retention in early years), suggesting that early deficits tend to persist without intervention. Conversely, higher-performance clusters show greater variability, potentially reflecting engagement factors. This study validates the use of unsupervised learning on touchscreen data to uncover heterogeneous developmental paths. The identified profiles serve as scalable, data-driven proxies for cognitive growth, offering a foundation for early screening tools and personalized pediatric interventions. --- ### Can Users Specify Driving Speed? Bench2Drive-Speed: Benchmark and Baselines for Desired-Speed Conditioned Autonomous Driving Authors: Yuqian Shao, Xiaosong Jia, Langechuan Liu, Junchi Yan Published: 2026-03-26 Categories: cs.RO, cs.CV arXiv: https://arxiv.org/abs/2603.25672v1 End-to-end autonomous driving (E2E-AD) has achieved remarkable progress. However, one practical and useful function has been long overlooked: users may wish to customize the desired speed of the policy or specify whether to allow the autonomous vehicle to overtake. To bridge this gap, we present Bench2Drive-Speed, a benchmark with metrics, dataset, and baselines for desired-speed conditioned autonomous driving. We introduce explicit inputs of users' desired target-speed and overtake/follow instructions to driving policy models. We design quantitative metrics, including Speed-Adherence Score and Overtake Score, to measure how faithfully policies follow user specifications, while remaining compatible with standard autonomous driving metrics. To enable training of speed-conditioned policies, one approach is to collect expert demonstrations that strictly follow speed requirements, an expensive and unscalable process in the real world. An alternative is to adapt existing regular driving data by treating the speed observed in future frames as the target speed for training. To investigate this, we construct CustomizedSpeedDataset, composed of 2,100 clips annotated with experts demonstrations, enabling systematic investigation of supervision strategies. Our experiments show that, under proper re-annotation, models trained on regular driving data perform comparably to on expert demonstrations, suggesting that speed supervision can be introduced without additional complex real-world data collection. Furthermore, we find that while target-speed following can be achieved without degrading regular driving performance, executing overtaking commands remains challenging due to the inherent difficulty of interactive behaviors. All code, datasets and baselines are available at https://github.com/Thinklab-SJTU/Bench2Drive-Speed --- ### Uncertainty-Guided Label Rebalancing for CPS Safety Monitoring Authors: John Ayotunde, Qinghua Xu, Guancheng Wang, Lionel C. Briand Published: 2026-03-26 Categories: cs.LG, cs.SE arXiv: https://arxiv.org/abs/2603.25670v4 Safety monitoring is essential for Cyber-Physical Systems (CPSs). However, unsafe events are rare in real-world CPS operations, creating an extreme class imbalance that degrades safety predictors. Standard rebalancing techniques perform poorly on time-series CPS telemetry, either generating unrealistic synthetic samples or overfitting on the minority class. Meanwhile, behavioral uncertainty in CPS operations, defined as the degree of doubt or uncertainty in CPS decisions , is often correlated with safety outcomes but unexplored in safety monitoring. To that end, we propose U-Balance, a supervised approach that leverages behavioral uncertainty to rebalance imbalanced datasets prior to training a safety predictor. U-Balance first trains a GatedMLP-based uncertainty predictor that summarizes each telemetry window into distributional kinematic features and outputs an uncertainty score. It then applies an uncertainty-guided label rebalancing (uLNR) mechanism that probabilistically relabels $\textit{safe}$-labeled windows with unusually high uncertainty as $\textit{unsafe}$, thereby enriching the minority class with informative boundary samples without synthesizing new data. Finally, a safety predictor is trained on the rebalanced dataset for safety monitoring. We evaluate U-Balance on a large-scale UAV benchmark with a 46:1 safe-to-unsafe ratio. Results confirm a moderate but significant correlation between behavioral uncertainty and safety. We then identify uLNR as the most effective strategy to exploit uncertainty information, compared to direct early and late fusion. U-Balance achieves a 0.806 F1 score, outperforming the strongest baseline by 14.3 percentage points, while maintaining competitive inference efficiency. Ablation studies confirm that both the GatedMLP-based uncertainty predictor and the uLNR mechanism contribute significantly to U-Balance's effectiveness. --- ### Fast-dVLA: Accelerating Discrete Diffusion VLA to Real-Time Performance Authors: Wenxuan Song, Jiayi Chen, Shuai Chen, Jingbo Wang, Pengxiang Ding, Han Zhao, Yikai Qin, Xinhu Zheng, Donglin Wang, Yan Wang, Haoang Li Published: 2026-03-26 Categories: cs.RO, cs.CV arXiv: https://arxiv.org/abs/2603.25661v3 This paper proposes a novel approach to address the challenge that pretrained VLA models often fail to effectively improve performance and reduce adaptation costs during standard supervised finetuning (SFT). Some advanced finetuning methods with auxiliary training objectives can improve performance and reduce the number of convergence steps. However, they typically incur significant computational overhead due to the additional losses from auxiliary tasks. To simultaneously achieve the enhanced capabilities of auxiliary training with the simplicity of standard SFT, we decouple the two objectives of auxiliary task training within the parameter space, namely, enhancing general capabilities and fitting task-specific action distributions. To deliver this goal, we only need to train the model to converge on a small-scale task set using two distinct training strategies. The difference between the resulting model parameters can then be interpreted as capability vectors provided by auxiliary tasks. These vectors are then merged with pretrained parameters to form a capability-enhanced meta model. Moreover, when standard SFT is augmented with a lightweight orthogonal regularization loss, the merged model attains performance comparable to auxiliary finetuned baselines with reduced computational overhead. Experimental results demonstrate that this approach is highly effective across diverse robot tasks. Project page: https://chris1220313648.github.io/Fast-dVLA/ --- ### Instance-optimal stochastic convex optimization: Can we improve upon sample-average and robust stochastic approximation? Authors: Liwei Jiang, Ashwin Pananjady Published: 2026-03-26 Categories: math.OC, stat.ML arXiv: https://arxiv.org/abs/2603.25657v1 We study the unconstrained minimization of a smooth and strongly convex population loss function under a stochastic oracle that introduces both additive and multiplicative noise; this is a canonical and widely-studied setting that arises across operations research, signal processing, and machine learning. We begin by showing that standard approaches such as sample average approximation and robust (or averaged) stochastic approximation can lead to suboptimal -- and in some cases arbitrarily poor -- performance with realistic finite sample sizes. In contrast, we demonstrate that a carefully designed variance reduction strategy, which we term VISOR for short, can significantly outperform these approaches while using the same sample size. Our upper bounds are complemented by finite-sample, information-theoretic local minimax lower bounds, which highlight fundamental, instance-dependent factors that govern the performance of any estimator. Taken together, these results demonstrate that an accelerated variant of VISOR is instance-optimal, achieving the best possible sample complexity up to logarithmic factors while also attaining optimal oracle complexity. We apply our theory to generalized linear models and improve upon classical results. In particular, we obtain the best-known non-asymptotic, instance-dependent generalization error bounds for stochastic methods, even in linear regression. ---