# Research Papers — 2026-03 · bundle page 116 of 243 > Full-text markdown of papers #5751–5800 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/116 [← Previous bundle](https://aipod.dev/papers/archive/2026-03/page/115.md) [Next bundle →](https://aipod.dev/papers/archive/2026-03/page/117.md) ### ASDA: Automated Skill Distillation and Adaptation for Financial Reasoning Authors: Tik Yu Yim, Wenting Tan, Sum Yee Chan, Tak-Wah Lam, Siu Ming Yiu Published: 2026-03-17 Categories: cs.CL, cs.AI, cs.CE arXiv: https://arxiv.org/abs/2603.16112v1 Adapting large language models (LLMs) to specialized financial reasoning typically requires expensive fine-tuning that produces model-locked expertise. Training-free alternatives have emerged, yet our experiments show that leading methods (GEPA and ACE) achieve only marginal gains on the FAMMA financial reasoning benchmark, exposing the limits of unstructured text optimization for complex, multi-step domain reasoning. We introduce Automated Skill Distillation and Adaptation (ASDA), a framework that automatically generates structured skill artifacts through iterative error-corrective learning without modifying model weights. A teacher model analyzes a student model's failures on financial reasoning tasks, clusters errors by subfield and error type, and synthesizes skill files containing reasoning procedures, code templates, and worked examples, which are dynamically injected during inference. Evaluated on FAMMA, ASDA achieves up to +17.33% improvement on arithmetic reasoning and +5.95% on non-arithmetic reasoning, substantially outperforming all training-free baselines. The resulting skill artifacts are human-readable, version-controlled, and compatible with the Agent Skills open standard, offering any organization with a labeled domain dataset a practical and auditable path to domain adaptation without weight access or retraining. --- ### VIGIL: Towards Edge-Extended Agentic AI for Enterprise IT Support Authors: Sarthak Ahuja, Neda Kordjazi, Evren Yortucboylu, Vishaal Kapoor, Mariam Dundua, Yiming Li, Derek Ho, Vaibhavi Padala, Jennifer Whitted, Rebecca Steinert Published: 2026-03-17 Categories: cs.AI arXiv: https://arxiv.org/abs/2603.16110v1 Enterprise IT support is constrained by heterogeneous devices, evolving policies, and long-tail failure modes that are difficult to resolve centrally. We present VIGIL, an edge-extended agentic AI system that deploys desktop-resident agents to perform situated diagnosis, retrieval over enterprise knowledge, and policy-governed remediation directly on user devices with explicit consent and end-to-end observability. In a 10-week pilot of VIGIL's operational loop on 100 resource-constrained endpoints, VIGIL reduces interaction rounds by 39%, achieves at least 4 times faster diagnosis, and supports self-service resolution in 82% of matched cases. Users report excellent usability, high trust, and low cognitive workload across four validated instruments, with qualitative feedback highlighting transparency as critical for trust. Notably, users rated the system higher when no historical matches were available, suggesting on-device diagnosis provides value independent of knowledge base coverage. This pilot establishes safety and observability foundations for fleet-wide continuous improvement. --- ### RepoReviewer: A Local-First Multi-Agent Architecture for Repository-Level Code Review Authors: Peng Zhang Published: 2026-03-17 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2603.16107v1 Repository-level code review requires reasoning over project structure, repository context, and file-level implementation details. Existing automated review workflows often collapse these tasks into a single pass, which can reduce relevance, increase duplication, and weaken prioritization. We present RepoReviewer, a local-first multi-agent system for automated GitHub repository review with a Python CLI, FastAPI API, LangGraph orchestration layer, and Next.js user interface. RepoReviewer decomposes review into repository acquisition, context synthesis, file-level analysis, finding prioritization, and summary generation. We describe the system design, implementation tradeoffs, developer-facing interfaces, and practical failure modes. Rather than claiming benchmark superiority, we frame RepoReviewer as a technical systems contribution: a pragmatic architecture for repository-level automated review, accompanied by reusable evaluation and reporting infrastructure for future empirical study. --- ### Frequency Matters: Fast Model-Agnostic Data Curation for Pruning and Quantization Authors: Francesco Pio Monaco, Elia Cunegatti, Flavio Vella, Giovanni Iacca Published: 2026-03-17 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2603.16105v4 Post-training model compression is essential for enhancing the portability of Large Language Models (LLMs) while preserving their performance. While several compression approaches have been proposed, less emphasis has been placed on selecting the most suitable set of data (the so-called \emph{calibration data}) for finding the compressed model configuration. The choice of calibration data is a critical step in preserving model capabilities both intra- and inter-tasks. In this work, we address the challenge of identifying high-performance calibration sets for both pruning and quantization by analyzing intrinsic data properties rather than model-specific signals. We introduce ZipCal, a model-agnostic data curation strategy that maximizes lexical diversity based on Zipfian power laws. Experiments demonstrate that our method outperforms standard uniform random sampling across various pruning benchmarks. Notably, it also performs on par, in terms of downstream performance, with a state-of-the-art method that relies on model perplexity. The latter becomes prohibitively expensive for large-scale models and datasets, while ZipCal is on average $\sim$240$\times$ faster due to its tractable linear complexity. We make the code and the experiments available at https://github.com/FrancescoMonaco/ZipCal. --- ### Frequency Matters: Fast Model-Agnostic Data Curation for Pruning and Quantization Authors: Francesco Pio Monaco, Elia Cunegatti, Flavio Vella, Giovanni Iacca Published: 2026-03-17 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2603.16105v3 Post-training model compression is essential for enhancing the portability of Large Language Models (LLMs) while preserving their performance. While several compression approaches have been proposed, less emphasis has been placed on selecting the most suitable set of data (the so-called \emph{calibration data}) for finding the compressed model configuration. The choice of calibration data is a critical step in preserving model capabilities both intra- and inter-tasks. In this work, we address the challenge of identifying high-performance calibration sets for both pruning and quantization by analyzing intrinsic data properties rather than model-specific signals. We introduce \texttt{\textbf{ZipCal}}, a model-agnostic data curation strategy that maximizes lexical diversity based on Zipfian power laws. Experiments demonstrate that our method consistently outperforms standard uniform random sampling across various pruning benchmarks. Notably, it also performs on par, in terms of downstream performance, with a state-of-the-art method that relies on model perplexity. The latter becomes prohibitively expensive at large-scale models and datasets, while \texttt{\textbf{ZipCal}} is on average $\sim$240$\times$ faster due to its tractable linear complexity\footnote{We make the code and the experiments available at https://github.com/FrancescoMonaco/ZipCal.}. --- ### Efficient LLM Serving for Agentic Workflows: A Data Systems Perspective Authors: Noppanat Wadlom, Junyi Shen, Yao Lu Published: 2026-03-17 Categories: cs.MA, cs.AI, cs.DB arXiv: https://arxiv.org/abs/2603.16104v1 Agentic workflows are composed of sequences of interdependent Large Language Model (LLM) calls, and they have become a dominant workload in modern AI systems. These workflows exhibit extensive redundancy from overlapping prompts and intermediate results due to speculative and parallel exploration. Existing LLM serving systems, such as vLLM, focus on optimizing individual inference calls and overlook cross-call dependencies, leading to significant inefficiencies. This paper rethinks LLM and agent serving from a data systems perspective and introduces Helium, a workflow-aware serving framework that models agentic workloads as query plans and treats LLM invocations as first-class operators. Helium integrates proactive caching and cache-aware scheduling to maximize reuse across prompts, KV states, and workflows. Through these techniques, Helium bridges classic query optimization principles with LLM serving, achieving up to 1.56x speedup over state-of-the-art agent serving systems on various workloads. Our results demonstrate that end-to-end optimization across workflows is essential for scalable and efficient LLM-based agents. --- ### NanoGS: Training-Free Gaussian Splat Simplification Authors: Butian Xiong, Rong Liu, Tiantian Zhou, Meida Chen, Zhiwen Fan, Andrew Feng Published: 2026-03-17 Categories: cs.CV, cs.GR arXiv: https://arxiv.org/abs/2603.16103v3 3D Gaussian Splat (3DGS) enables high-fidelity, real-time novel view synthesis by representing scenes with large sets of anisotropic primitives, but often requires millions of Splats, incurring significant storage and transmission costs. Most existing compression methods rely on GPU-intensive post-training optimization with calibrated images, limiting practical deployment. We introduce \textbf{NanoGS}, a training-free and lightweight framework for Gaussian Splat simplification. Instead of relying on image-based rendering supervision, NanoGS formulates simplification as local pairwise merging over a sparse spatial graph. The method approximates a pair of Gaussians with a single primitive using mass preserved moment matching and evaluates merge quality through a principled merge cost between the original mixture and its approximation. By restricting merge candidates to local neighborhoods and selecting compatible pairs efficiently, NanoGS produces compact Gaussian representations while preserving scene structure and appearance. NanoGS operates directly on existing Gaussian Splat models, runs efficiently on CPU, and preserves the standard 3DGS parameterization, enabling seamless integration with existing rendering pipelines. Experiments demonstrate that NanoGS substantially reduces primitive count while maintaining high rendering fidelity, providing an efficient and practical solution for Gaussian Splat simplification. Our project website is available at \href{https://saliteta.github.io/NanoGS/}{https://saliteta.github.io/NanoGS/}. --- ### NanoGS: Training-Free Gaussian Splat Simplification Authors: Butian Xiong, Rong Liu, Tiantian Zhou, Meida Chen, Zhiwen Fan, Andrew Feng Published: 2026-03-17 Categories: cs.CV, cs.GR arXiv: https://arxiv.org/abs/2603.16103v2 3D Gaussian Splat (3DGS) enables high-fidelity, real-time novel view synthesis by representing scenes with large sets of anisotropic primitives, but often requires millions of Splats, incurring significant storage and transmission costs. Most existing compression methods rely on GPU-intensive post-training optimization with calibrated images, limiting practical deployment. We introduce \textbf{NanoGS}, a training-free and lightweight framework for Gaussian Splat simplification. Instead of relying on image-based rendering supervision, NanoGS formulates simplification as local pairwise merging over a sparse spatial graph. The method approximates a pair of Gaussians with a single primitive using mass preserved moment matching and evaluates merge quality through a principled merge cost between the original mixture and its approximation. By restricting merge candidates to local neighborhoods and selecting compatible pairs efficiently, NanoGS produces compact Gaussian representations while preserving scene structure and appearance. NanoGS operates directly on existing Gaussian Splat models, runs efficiently on CPU, and preserves the standard 3DGS parameterization, enabling seamless integration with existing rendering pipelines. Experiments demonstrate that NanoGS substantially reduces primitive count while maintaining high rendering fidelity, providing an efficient and practical solution for Gaussian Splat simplification. Our project website is available at \href{https://saliteta.github.io/NanoGS/}{https://saliteta.github.io/NanoGS/}. --- ### Reevaluating the Intra-Modal Misalignment Hypothesis in CLIP Authors: Jonas Herzog, Yue Wang Published: 2026-03-17 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.16100v2 Recent research suggested that the embeddings produced by CLIP-like contrastive language-image training are suboptimal for image-only tasks. The main theory is that the inter-modal (language-image) alignment loss ignores intra-modal (image-image) alignment, leading to poorly calibrated distances between images. In this study, we question this intra-modal misalignment hypothesis. We reexamine its foundational theoretical argument, the indicators used to support it, and the performance metrics affected. For the theoretical argument, we demonstrate that there are no such supposed degrees of freedom for image embedding distances. For the empirical measures, our findings reveal they yield similar results for language-image trained models (CLIP, SigLIP) and image-image trained models (DINO, SigLIP2). This indicates the observed phenomena do not stem from a misalignment specific to the former. Experiments on the commonly studied intra-modal tasks retrieval and few-shot classification confirm that addressing task ambiguity, not supposed misalignment, is key for best results. --- ### OneWorld: Taming Scene Generation with 3D Unified Representation Autoencoder Authors: Sensen Gao, Zhaoqing Wang, Qihang Cao, Dongdong Yu, Changhu Wang, Tongliang Liu, Mingming Gong, Jiawang Bian Published: 2026-03-17 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.16099v1 Existing diffusion-based 3D scene generation methods primarily operate in 2D image/video latent spaces, which makes maintaining cross-view appearance and geometric consistency inherently challenging. To bridge this gap, we present OneWorld, a framework that performs diffusion directly within a coherent 3D representation space. Central to our approach is the 3D Unified Representation Autoencoder (3D-URAE); it leverages pretrained 3D foundation models and augments their geometry-centric nature by injecting appearance and distilling semantics into a unified 3D latent space. Furthermore, we introduce token-level Cross-View-Correspondence (CVC) consistency loss to explicitly enforce structural alignment across views, and propose Manifold-Drift Forcing (MDF) to mitigate train-inference exposure bias and shape a robust 3D manifold by mixing drifted and original representations. Comprehensive experiments demonstrate that OneWorld generates high-quality 3D scenes with superior cross-view consistency compared to state-of-the-art 2D-based methods. Our code will be available at https://github.com/SensenGao/OneWorld. --- ### LICA: Layered Image Composition Annotations for Graphic Design Research Authors: Elad Hirsch, Shubham Yadav, Mohit Garg, Purvanshi Mehta Published: 2026-03-17 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2603.16098v2 We introduce LICA (Layered Image Composition Annotations), a large scale dataset of 1,550,244 multi-layer graphic design compositions designed to advance structured understanding and generation of graphic layouts. In addition to rendered PNG images, LICA represents each design as a hierarchical composition of typed components including text, image, vector, and group elements, each paired with rich per-element metadata such as spatial geometry, typographic attributes, opacity, and visibility. The dataset spans 20 design categories and 971,850 unique templates, providing broad coverage of real-world design structures. We further introduce graphic design video as a new and largely unexplored challenge for current vision-language models through 27,261 animated layouts annotated with per-component keyframes and motion parameters. Beyond scale, LICA establishes a new paradigm of research tasks for graphic design, enabling structured investigations into problems such as layer-aware inpainting, structured layout generation, controlled design editing, and temporally-aware generative modeling. By representing design as a system of compositional layers and relationships, the dataset supports research on models that operate directly on design structure rather than pixels alone. --- ### Diffusion Models for Joint Audio-Video Generation Authors: Alejandro Paredes La Torre Published: 2026-03-17 Categories: cs.SD, cs.AI, cs.CV, cs.MM arXiv: https://arxiv.org/abs/2603.16093v1 Multimodal generative models have shown remarkable progress in single-modality video and audio synthesis, yet truly joint audio-video generation remains an open challenge. In this paper, I explore four key contributions to advance this field. First, I release two high-quality, paired audio-video datasets. The datasets consisting on 13 hours of video-game clips and 64 hours of concert performances, each segmented into consistent 34-second samples to facilitate reproducible research. Second, I train the MM-Diffusion architecture from scratch on our datasets, demonstrating its ability to produce semantically coherent audio-video pairs and quantitatively evaluating alignment on rapid actions and musical cues. Third, I investigate joint latent diffusion by leveraging pretrained video and audio encoder-decoders, uncovering challenges and inconsistencies in the multimodal decoding stage. Finally, I propose a sequential two-step text-to-audio-video generation pipeline: first generating video, then conditioning on both the video output and the original prompt to synthesize temporally synchronized audio. My experiments show that this modular approach yields high-fidelity generations of audio video generation. --- ### Parallel In-context Learning for Large Vision Language Models Authors: Shin'ya Yamaguchi, Daiki Chijiwa, Tamao Sakao, Taku Hasegawa Published: 2026-03-17 Categories: cs.CV, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2603.16092v1 Large vision-language models (LVLMs) employ multi-modal in-context learning (MM-ICL) to adapt to new tasks by leveraging demonstration examples. While increasing the number of demonstrations boosts performance, they incur significant inference latency due to the quadratic computational cost of Transformer attention with respect to the context length. To address this trade-off, we propose Parallel In-Context Learning (Parallel-ICL), a plug-and-play inference algorithm. Parallel-ICL partitions the long demonstration context into multiple shorter, manageable chunks. It processes these chunks in parallel and integrates their predictions at the logit level, using a weighted Product-of-Experts (PoE) ensemble to approximate the full-context output. Guided by ensemble learning theory, we introduce principled strategies for Parallel-ICL: (i) clustering-based context chunking to maximize inter-chunk diversity and (ii) similarity-based context compilation to weight predictions by query relevance. Extensive experiments on VQA, image captioning, and classification benchmarks demonstrate that Parallel-ICL achieves performance comparable to full-context MM-ICL, while significantly improving inference speed. Our work offers an effective solution to the accuracy-efficiency trade-off in MM-ICL, enabling dynamic task adaptation with substantially reduced inference overhead. --- ### CounterRefine: Answer-Conditioned Counterevidence Retrieval for Inference-Time Knowledge Repair in Factual Question Answering Authors: Tianyi Huang, Ying Kai Deng Published: 2026-03-17 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2603.16091v3 In factual question answering, many errors are not failures of access but failures of commitment: the system retrieves relevant evidence, yet still settles on the wrong answer. We present CounterRefine, a lightweight repair layer for short-form RAG that treats the first answer as a hypothesis to test. Given a draft, CounterRefine issues answer-conditioned expansion queries to retrieve candidate-specific evidence, then applies a constrained KEEP or REVISE refinement step whose proposed revisions are accepted only after deterministic validation. The design is intentionally narrow: it adds one evidence-gathering pass and one guarded refinement call rather than replacing the retriever or building a broad agentic system. On the full SimpleQA benchmark, CounterRefine improves a matched one-pass RAG baseline by up to 5.8 correct-rate points; in the full Claude trace, it changes only 5.6% of outputs, with 180 beneficial outcome changes and 8 harmful ones. These findings suggest a simple but important direction for knowledgeable foundation models: beyond accessing evidence, they should also be able to use that evidence to reconsider and, when necessary, repair their own answers. --- ### RecBundle: A Next-Generation Geometric Paradigm for Explainable Recommender Systems Authors: Hui Wang, Tianzhu Hu, Mingming Li, Xi Zhou, Chun Gan, Jiao Dai, Jizhong Han, Songlin Hu, Tao Guo Published: 2026-03-17 Categories: cs.IR, cs.AI arXiv: https://arxiv.org/abs/2603.16088v1 Recommender systems are inherently dynamic feedback loops where prolonged local interactions accumulate into macroscopic structural degradation such as information cocoons. Existing representation learning paradigms are universally constrained by the assumption of a single flat space, forcing topologically grounded user associations and semantically driven historical interactions to be fitted within the same vector space. This excessive coupling of heterogeneous information renders it impossible for researchers to mechanistically distinguish and identify the sources of systemic bias. To overcome this theoretical bottleneck, we introduce Fiber Bundle from modern differential geometry and propose a novel geometric analysis paradigm for recommender systems. This theory naturally decouples the system space into two hierarchical layers: the base manifold formed by user interaction networks, and the fibers attached to individual user nodes that carry their dynamic preferences. Building upon this, we construct RecBundle, a framework oriented toward next-generation recommender systems that formalizes user collaboration as geometric connection and parallel transport on the base manifold, while mapping content evolution to holonomy transformations on fibers. From this foundation, we identify future application directions encompassing quantitative mechanisms for information cocoons and evolutionary bias, geometric meta-theory for adaptive recommendation, and novel inference architectures integrating large language models (LLMs). Empirical analysis on real-world MovieLens and Amazon Beauty datasets validates the effectiveness of this geometric framework. --- ### Towards the Vision-Sound-Language-Action Paradigm: The HEAR Framework for Sound-Centric Manipulation Authors: Chang Nie, Tianchen Deng, Guangming Wang, Zhe Liu, Hesheng Wang Published: 2026-03-17 Categories: cs.RO, cs.AI, cs.CV, cs.SD arXiv: https://arxiv.org/abs/2603.16086v2 While recent Vision-Language-Action (VLA) models have begun to incorporate audio, they typically treat sound as static pre-execution prompts or focus exclusively on human speech. This leaves a significant gap in real-time, sound-centric manipulation where fleeting environmental acoustics provide critical state verification during task execution. Consequently, key sounds are easily missed due to low-frequency updates or system latency. This problem is exacerbated by action chunking with open-loop execution, which creates a Blind Execution Interval where acoustic events are lost between discrete audio observation windows. Recognizing the necessity of continuous auditory awareness, we formalize Vision-Sound-Language-Action (VSLA) as a continuous control paradigm conditioned on vision, streaming audio, language, and proprioception under delayed decision loops. As an instantiation, we introduce HEAR, a VSLA framework integrating four components: (i) a streaming Historizer to maintain a compact, causal audio context across execution gaps; (ii) an Envisioner adapted from omni foundation models to reason over multi-sensory inputs; (iii) an Advancer, formulated as an audio world model, to learn temporal dynamics by predicting near-future audio codes; and (iv) a flow-matching Realizer policy to generate smooth action chunks. To address the scarcity of pretraining data and evaluations for VSLA, we construct OpenX-Sound for pretraining, alongside HEAR-Bench, the first sound-centric manipulation benchmark with strict causal timing rules. Our results suggest that robust sound-centric manipulation necessitates causal persistence and explicit temporal learning. This framework provides a practical step toward multi-sensory foundation models for embodied agents, enabling robots to perceive and interact with dynamic environments. Code and videos are available at https://hear.irmv.top --- ### Towards the Vision-Sound-Language-Action Paradigm: The HEAR Framework for Sound-Centric Manipulation Authors: Chang Nie, Tianchen Deng, Guangming Wang, Zhe Liu, Hesheng Wang Published: 2026-03-17 Categories: cs.RO, cs.AI, cs.CV, cs.SD arXiv: https://arxiv.org/abs/2603.16086v1 While recent Vision-Language-Action (VLA) models have begun to incorporate audio, they typically treat sound as static pre-execution prompts or focus exclusively on human speech. This leaves a significant gap in real-time, sound-centric manipulation where fleeting environmental acoustics provide critical state verification during task execution. Consequently, key sounds are easily missed due to low-frequency updates or system latency. This problem is exacerbated by action chunking with open-loop execution, which creates a Blind Execution Interval where acoustic events are lost between discrete audio observation windows. Recognizing the necessity of continuous auditory awareness, we formalize Vision-Sound-Language-Action (VSLA) as a continuous control paradigm conditioned on vision, streaming audio, language, and proprioception under delayed decision loops. As an instantiation, we introduce HEAR, a VSLA framework integrating four components: (i) a streaming Historizer to maintain a compact, causal audio context across execution gaps; (ii) an Envisioner adapted from omni foundation models to reason over multi-sensory inputs; (iii) an Advancer, formulated as an audio world model, to learn temporal dynamics by predicting near-future audio codes; and (iv) a flow-matching Realizer policy to generate smooth action chunks. To address the scarcity of pretraining data and evaluations for VSLA, we construct OpenX-Sound for pretraining, alongside HEAR-Bench, the first sound-centric manipulation benchmark with strict causal timing rules. Our results suggest that robust sound-centric manipulation necessitates causal persistence and explicit temporal learning. This framework provides a practical step toward multi-sensory foundation models for embodied agents, enabling robots to perceive and interact with dynamic environments. Code and videos are available at https://hear.irmv.top. --- ### Interact3D: Compositional 3D Generation of Interactive Objects Authors: Hui Shan, Keyang Luo, Ming Li, Sizhe Zheng, Yanwei Fu, Zhen Chen, Xiangru Huang Published: 2026-03-17 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2603.16085v2 Recent breakthroughs in 3D generation have enabled the synthesis of high-fidelity individual assets. However, generating 3D compositional objects from single images--particularly under occlusions--remains challenging. Existing methods often degrade geometric details in hidden regions and fail to preserve the underlying object-object spatial relationships (OOR). We present a novel framework Interact3D designed to generate physically plausible interacting 3D compositional objects. Our approach first leverages advanced generative priors to curate high-quality individual assets with a unified 3D guidance scene. To physically compose these assets, we then introduce a robust two-stage composition pipeline. Based on the 3D guidance scene, the primary object is anchored through precise global-to-local geometric alignment (registration), while subsequent geometries are integrated using a differentiable Signed Distance Field (SDF)-based optimization that explicitly penalizes geometry intersections. To reduce challenging collisions, we further deploy a closed-loop, agentic refinement strategy. A Vision-Language Model (VLM) autonomously analyzes multi-view renderings of the composed scene, formulates targeted corrective prompts, and guides an image editing module to iteratively self-correct the generation pipeline. Extensive experiments demonstrate that Interact3D successfully produces promising collsion-aware compositions with improved geometric fidelity and consistent spatial relationships. --- ### Structured prototype regularization for synthetic-to-real driving scene parsing Authors: Jiahe Fan, Xiao Ma, Sergey Vityazev, George Giakos, Shaolong Shu, Rui Fan Published: 2026-03-17 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.16083v1 Driving scene parsing is critical for autonomous vehicles to operate reliably in complex real-world traffic environments. To reduce the reliance on costly pixel-level annotations, synthetic datasets with automatically generated labels have become a popular alternative. However, models trained on synthetic data often perform poorly when applied to real-world scenes due to the synthetic-to-real domain gap. Despite the success of unsupervised domain adaptation in narrowing this gap, most existing methods mainly focus on global feature alignment while overlooking the semantic structure of the feature space. As a result, semantic relations among classes are insufficiently modeled, limiting the model's ability to generalize. To address these challenges, this study introduces a novel unsupervised domain adaptation framework that explicitly regularizes semantic feature structures to significantly enhance driving scene parsing performance in real-world scenarios. Specifically, the proposed method enforces inter-class separation and intra-class compactness by leveraging class-specific prototypes, thereby enhancing the discriminability and structural coherence of feature clusters. An entropy-based noise filtering strategy improves the reliability of pseudo labels, while a pixel-level attention mechanism further refines feature alignment. Extensive experiments on representative benchmarks demonstrate that the proposed method consistently outperforms recent state-of-the-art methods. These results underscore the importance of preserving semantic structure for robust synthetic-to-real adaptation in driving scene parsing tasks. --- ### A Depth-Aware Comparative Study of Euclidean and Hyperbolic Graph Neural Networks on Bitcoin Transaction Systems Authors: Ankit Ghimire, Saydul Akbar Murad, Nick Rahimi Published: 2026-03-17 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.16080v1 Bitcoin transaction networks are large scale socio- technical systems in which activities are represented through multi-hop interaction patterns. Graph Neural Networks(GNNs) have become a widely adopted tool for analyzing such systems, supporting tasks such as entity detection and transaction classification. Large-scale datasets like Elliptic have allowed for a rise in the analysis of these systems and in tasks such as fraud detection. In these settings, the amount of transactional context available to each node is determined by the neighborhood aggregation and sampling strategies, yet the interaction between these receptive fields and embedding geometry has received limited attention. In this work, we conduct a controlled comparison of Euclidean and tangent-space hyperbolic GNNs for node classification on a large Bitcoin transaction graph. By explicitly varying the neighborhood while keeping the model architecture and dimensionality fixed, we analyze the differences in two embedding spaces. We further examine optimization behavior and observe that joint selection of learning rate and curvature plays a critical role in stabilizing high-dimensional hyperbolic embeddings. Overall, our findings provide practical insights into the role of embedding geometry and neighborhood depth when modeling large-scale transaction networks, informing the deployment of hyperbolic GNNs for computational social systems. --- ### Volumetrically Consistent Implicit Atlas Learning via Neural Diffeomorphic Flow for Placenta MRI Authors: Athena Taymourtash, S. Mazdak Abulnaga, Esra Abaci Turk, P. Ellen Grant, Polina Golland Published: 2026-03-17 Categories: cs.CV, cs.GR arXiv: https://arxiv.org/abs/2603.16078v1 Establishing dense volumetric correspondences across anatomical shapes is essential for group-level analysis but remains challenging for implicit neural representations. Most existing implicit registration methods rely on supervision near the zero-level set and thus capture only surface correspondences, leaving interior deformations under-constrained. We introduce a volumetrically consistent implicit model that couples reconstruction of signed distance functions (SDFs) with neural diffeomorphic flow to learn a shared canonical template of the placenta. Volumetric regularization, including Jacobian-determinant and biharmonic penalties, suppresses local folding and promotes globally coherent deformations. In the motivating application to placenta MRI, our formulation jointly reconstructs individual placentas, aligns them to a population-derived implicit template, and enables voxel-wise intensity mapping in a unified canonical space. Experiments on in-vivo placenta MRI scans demonstrate improved geometric fidelity and volumetric alignment over surface-based implicit baseline methods, yielding anatomically interpretable and topologically consistent flattening suitable for group analysis. --- ### MDM-Prime-v2: Binary Encoding and Index Shuffling Enable Scaling of Diffusion Language Models Authors: Chen-Hao Chao, Wei-Fang Sun, Junwei Quan, Chun-Yi Lee, Rahul G. Krishnan Published: 2026-03-17 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.16077v4 Masked diffusion models (MDM) exhibit superior generalization when learned using a Partial masking scheme (Prime). This approach converts tokens into sub-tokens and models the diffusion process at the sub-token level. We identify two limitations of the MDM-Prime framework. First, we find that the functional form of the subtokenizer significantly increases the cross-entropy loss in the objective when paired with commonly used Byte-Pair-Encoding (BPE) tokenizers. Second, we lack tools to guide the hyperparameter choice of the token granularity in the subtokenizer. To address these limitations, we analyze the optimal design of the subtokenizer that minimizes MDM-Prime training objective and develop MDM-Prime-v2, a masked diffusion language model which incorporates Binary Encoding and Index Shuffling. Our analysis characterizes how token granularity and sub-token entropy influence the training objective and downstream performance, providing principled criteria for subtokenizer design. When extending the model size to 1.1B parameters, MDM-Prime-v2 demonstrates superior average zero-shot accuracy across eight commonsense reasoning benchmarks, outperforming similar-sized baselines including GPT-Neo, OPT, Pythia, Bloom, SMDM, and TinyLLaMA. --- ### MDM-Prime-v2: Binary Encoding and Index Shuffling Enable Scaling of Diffusion Language Models Authors: Chen-Hao Chao, Wei-Fang Sun, Junwei Quan, Chun-Yi Lee, Rahul G. Krishnan Published: 2026-03-17 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.16077v3 Masked diffusion models (MDM) exhibit superior generalization when learned using a Partial masking scheme (Prime). This approach converts tokens into sub-tokens and models the diffusion process at the sub-token level. We identify two limitations of the MDM-Prime framework. First, we find that the functional form of the subtokenizer significantly increases the cross-entropy loss in the objective when paired with commonly used Byte-Pair-Encoding (BPE) tokenizers. Second, we lack tools to guide the hyperparameter choice of the token granularity in the subtokenizer. To address these limitations, we analyze the optimal design of the subtokenizer that minimizes MDM-Prime training objective and develop MDM-Prime-v2, a masked diffusion language model which incorporates Binary Encoding and Index Shuffling. Our analysis characterizes how token granularity and sub-token entropy influence the training objective and downstream performance, providing principled criteria for subtokenizer design. When extending the model size to 1.1B parameters, MDM-Prime-v2 demonstrates superior average zero-shot accuracy across eight commonsense reasoning benchmarks, outperforming similar-sized baselines including GPT-Neo, OPT, Pythia, Bloom, SMDM, and TinyLLaMA. --- ### ClaimFlow: Tracing the Evolution of Scientific Claims in NLP Authors: Aniket Pramanick, Yufang Hou, Saif M. Mohammad, Iryna Gurevych Published: 2026-03-17 Categories: cs.CL arXiv: https://arxiv.org/abs/2603.16073v2 Scientific papers advance $\textit{claims}$ that later work supports, extends, or sometimes refutes. Yet existing methods for citation and claim analysis capture only fragments of this dialogue. In this work, we make these interactions explicit at the level of individual scientific claims. We introduce $\texttt{ClaimFlow}$, a claim-centric view of the NLP literature, built from $1{,}617$ ACL Anthology papers $(1979 - 2025)$ that are manually annotated with $5{,}689$ claims and $4{,}871$ cross-paper claim relations, indicating whether a citing paper $\texttt{supports}$, $\texttt{extends}$, $\texttt{qualifies}$, $\texttt{refutes}$, or references a cited claim as $\texttt{background}$. Building on $\texttt{ClaimFlow}$, we define a new task -- $\textit{Claim Relation Classification}$ -- which requires models to infer the scientific stance toward a cited claim from the text and citation context. Evaluating neural models and large language models on this task, we report baseline performance of $0.81$ macro-F1, suggesting that the task is tractable while leaving room for improvement. We then scale this framework to $\sim$$13k$ NLP papers to study claim evolution across decades of NLP research. We show that $63.5\%$ claims are never reused; only $11.1\%$ are ever challenged. Widely propagated claims are more often $\textit{reshaped}$ through qualification and extension than supported or refuted. Overall, $\texttt{ClaimFlow}$ offers a lens for examining how ideas shift and mature within NLP. --- ### SEAHateCheck: Functional Tests for Detecting Hate Speech in Low-Resource Languages of Southeast Asia Authors: Ri Chi Ng, Aditi Kumaresan, Yujia Hu, Roy Ka-Wei Lee Published: 2026-03-17 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2603.16070v1 Hate speech detection relies heavily on linguistic resources, which are primarily available in high-resource languages such as English and Chinese, creating barriers for researchers and platforms developing tools for low-resource languages in Southeast Asia, where diverse socio-linguistic contexts complicate online hate moderation. To address this, we introduce SEAHateCheck, a pioneering dataset tailored to Indonesia, Thailand, the Philippines, and Vietnam, covering Indonesian, Tagalog, Thai, and Vietnamese. Building on HateCheck's functional testing framework and refining SGHateCheck's methods, SEAHateCheck provides culturally relevant test cases, augmented by large language models and validated by local experts for accuracy. Experiments with state-of-the-art and multilingual models revealed limitations in detecting hate speech in specific low-resource languages. In particular, Tagalog test cases showed the lowest model accuracy, likely due to linguistic complexity and limited training data. In contrast, slang-based functional tests proved the hardest, as models struggled with culturally nuanced expressions. The diagnostic insights of SEAHateCheck further exposed model weaknesses in implicit hate detection and models' struggles with counter-speech expression. As the first functional test suite for these Southeast Asian languages, this work equips researchers with a robust benchmark, advancing the development of practical, culturally attuned hate speech detection tools for inclusive online content moderation. --- ### Resource Consumption Threats in Large Language Models Authors: Yuanhe Zhang, Xinyue Wang, Zhican Chen, Weiliu Wang, Zilu Zhang, Zhengshuo Gong, Zhenhong Zhou, Kun Wang, Li Sun, Yang Liu, Sen Su Published: 2026-03-17 Categories: cs.CR, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2603.16068v3 Given limited and costly computational infrastructure, resource efficiency is a key requirement for large language models (LLMs). Efficient LLMs increase service capacity for providers and reduce latency and API costs for users. Recent resource consumption threats induce excessive generation, degrading model efficiency and harming both service availability and economic sustainability. This survey presents a systematic review of threats to resource consumption in LLMs. We further establish a unified view of this emerging area by clarifying its scope and examining the problem along the full pipeline from threat induction to mechanism understanding and mitigation. Our goal is to clarify the problem landscape for this emerging area, thereby providing a clearer foundation for characterization and mitigation. --- ### Attribution Upsampling should Redistribute, Not Interpolate Authors: Vincenzo Buono, Peyman Sheikholharam Mashhadi, Mahmoud Rahat, Prayag Tiwari, Stefan Byttner Published: 2026-03-17 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2603.16067v1 Attribution methods in explainable AI rely on upsampling techniques that were designed for natural images, not saliency maps. Standard bilinear and bicubic interpolation systematically corrupts attribution signals through aliasing, ringing, and boundary bleeding, producing spurious high-importance regions that misrepresent model reasoning. We identify that the core issue is treating attribution upsampling as an interpolation problem that operates in isolation from the model's reasoning, rather than a mass redistribution problem where model-derived semantic boundaries must govern how importance flows. We present Universal Semantic-Aware Upsampling (USU), a principled method that reformulates upsampling through ratio-form mass redistribution operators, provably preserving attribution mass and relative importance ordering. Extending the axiomatic tradition of feature attribution to upsampling, we formalize four desiderata for faithful upsampling and prove that interpolation structurally violates three of them. These same three force any redistribution operator into a ratio form; the fourth selects the unique potential within this family, yielding USU. Controlled experiments on models with known attribution priors verify USU's formal guarantees; evaluation across ImageNet, CIFAR-10, and CUB-200 confirms consistent faithfulness improvements and qualitatively superior, semantically coherent explanations. --- ### Adaptive regularization parameter selection for high-dimensional inverse problems: A Bayesian approach with Tucker low-rank constraints Authors: Qing-Mei Yang, Da-Qing Zhang Published: 2026-03-17 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.16066v1 This paper introduces a novel variational Bayesian method that integrates Tucker decomposition for efficient high-dimensional inverse problem solving. The method reduces computational complexity by transforming variational inference from a high-dimensional space to a lower-dimensional core tensor space via Tucker decomposition. A key innovation is the introduction of per-mode precision parameters, enabling adaptive regularization for anisotropic structures. For instance, in directional image deblurring, learned parameters align with physical anisotropy, applying stronger regularization to critical directions (e.g., row vs. column axes). The method further estimates noise levels from data, eliminating reliance on prior knowledge of noise parameters (unlike conventional benchmarks such as the discrepancy principle (DP)). Experimental evaluations across 2D deblurring, 3D heat conduction, and Fredholm integral equations demonstrate consistent improvements in quantitative metrics (PSNR, SSIM) and qualitative visualizations (error maps, precision parameter trends) compared to L-curve criterion, generalized cross-validation (GCV), unbiased predictive risk estimator (UPRE), and DP. The approach scales to problems with 110,000 variables and outperforms existing methods by 0.73-2.09 dB in deblurring tasks and 6.75 dB in 3D heat conduction. Limitations include sensitivity to rank selection in Tucker decomposition and the need for theoretical analysis. Future work will explore automated rank selection and theoretical guarantees. This method bridges Bayesian theory and scalable computation, offering practical solutions for large-scale inverse problems in imaging, remote sensing, and scientific computing. --- ### Large Reward Models: Generalizable Online Robot Reward Generation with Vision-Language Models Authors: Yanru Wu, Weiduo Yuan, Ang Qi, Vitor Guizilini, Jiageng Mao, Yue Wang Published: 2026-03-17 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2603.16065v2 Reinforcement Learning (RL) has shown great potential in refining robotic manipulation policies, yet its efficacy remains strongly bottlenecked by the difficulty of designing generalizable reward functions. In this paper, we propose a framework for online policy refinement by adapting foundation VLMs into online reward generators. We develop a robust, scalable reward model based on a state-of-the-art VLM, trained on a large-scale, multi-source dataset encompassing real-world robot trajectories, human-object interactions, and diverse simulated environments. Unlike prior approaches that evaluate entire trajectories post-hoc, our method leverages the VLM to formulate a multifaceted reward signal comprising process, completion, and temporal contrastive rewards based on current visual observations. Initializing with a base policy trained via Imitation Learning (IL), we employ these VLM rewards to guide the model to correct sub-optimal behaviors in a closed-loop manner. We evaluate our framework on challenging long-horizon manipulation benchmarks requiring sequential execution and precise control. Crucially, our reward model operates in a purely zero-shot manner within these test environments. Experimental results demonstrate that our method significantly improves the success rate of the initial IL policy within just 30 RL iterations, demonstrating remarkable sample efficiency. This empirical evidence highlights that VLM-generated signals can provide reliable feedback to resolve execution errors, effectively eliminating the need for manual reward engineering and facilitating efficient online refinement for robot learning. --- ### ViT-AdaLA: Adapting Vision Transformers with Linear Attention Authors: Yifan Li, Seunghyun Yoon, Viet Dac Lai, Franck Dernoncourt, Jason Kuen, Yu Kong, Trung Bui Published: 2026-03-17 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.16063v1 Vision Transformers (ViTs) based vision foundation models (VFMs) have achieved remarkable performance across diverse vision tasks, but suffer from quadratic complexity that limits scalability to long sequences. Existing linear attention approaches for ViTs are typically trained from scratch, requiring substantial computational resources, while linearization-based methods developed for large language model decoders do not transfer well to ViTs. To address these challenges, we propose ViT-AdaLA, a novel framework for effectively adapting and transferring prior knowledge from VFMs to linear attention ViTs. ViT-AdaLA consists of three stages: attention alignment, feature alignment, and supervised fine-tuning. In the attention alignment stage, we align vanilla linear attention with the original softmax-based attention in each block to approximate the behavior of softmax attention. However, residual approximation errors inevitably accumulate across layers. We mitigate this by fine-tuning the linearized ViT to align its final-layer features with a frozen softmax VFM teacher. Finally, the adapted prior knowledge is transferred to downstream tasks through supervised fine-tuning. Extensive experiments on classification and segmentation tasks demonstrate the effectiveness and generality of ViT-AdaLA over various state-of-the-art linear attention counterpart. --- ### Safe Distributionally Robust Feature Selection under Covariate Shift Authors: Hiroyuki Hanada, Satoshi Akahane, Noriaki Hashimoto, Shion Takeno, Ichiro Takeuchi Published: 2026-03-17 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2603.16062v1 In practical machine learning, the environments encountered during the model development and deployment phases often differ, especially when a model is used by many users in diverse settings. Learning models that maintain reliable performance across plausible deployment environments is known as distributionally robust (DR) learning. In this work, we study the problem of distributionally robust feature selection (DRFS), with a particular focus on sparse sensing applications motivated by industrial needs. In practical multi-sensor systems, a shared subset of sensors is typically selected prior to deployment based on performance evaluations using many available sensors. At deployment, individual users may further adapt or fine-tune models to their specific environments. When deployment environments differ from those anticipated during development, this strategy can result in systems lacking sensors required for optimal performance. To address this issue, we propose safe-DRFS, a novel approach that extends safe screening from conventional sparse modeling settings to a DR setting under covariate shift. Our method identifies a feature subset that encompasses all subsets that may become optimal across a specified range of input distribution shifts, with finite-sample theoretical guarantees of no false feature elimination. --- ### ARISE: Agent Reasoning with Intrinsic Skill Evolution in Hierarchical Reinforcement Learning Authors: Yu Li, Rui Miao, Zhengling Qi, Tian Lan Published: 2026-03-17 Categories: cs.AI arXiv: https://arxiv.org/abs/2603.16060v2 The dominant paradigm for improving mathematical reasoning in language models relies on Reinforcement Learning with verifiable rewards. Yet existing methods treat each problem instance in isolation without leveraging the reusable strategies that emerge and accumulate during training. To this end, we introduce ARISE (Agent Reasoning via Intrinsic Skill Evolution), a hierarchical reinforcement learning framework, in which a shared policy operates both to manage skills at high-level and to generate responses at low-level (denoted as a Skills Manager and a Worker, respectively). The Manager maintains a tiered skill library through a dedicated skill generation rollout that performs structured summarization of successful solution traces (after execution), while employing a policy-driven selection mechanism to retrieve relevant skills to condition future rollouts (before execution). A hierarchical reward design guides the co-evolution of reasoning ability and library quality. Experiments on two base models and seven benchmarks spanning both competition mathematics and Omni-MATH show that ARISE consistently outperforms GRPO-family algorithms and memory-augmented baselines, with particularly notable gains on out-of-distribution tasks. Ablation studies confirm that each component contributes to the observed improvements and that library quality and reasoning performance improve in tandem throughout training. Code is available at \href{https://github.com/Skylanding/ARISE}{https://github.com/Skylanding/ARISE}. --- ### Population Annealing as a Discrete-Time Schrödinger Bridge Authors: Masayuki Ohzeki Published: 2026-03-17 Categories: cond-mat.stat-mech, quant-ph, stat.ML arXiv: https://arxiv.org/abs/2603.16056v1 We present a theoretical framework that reinterprets Population Annealing (PA) through the lens of the discrete-time Schrödinger Bridge (SB) problem. We demonstrate that the heuristic reweighting step in PA is derived by analytically solving the Schrödinger system without iterative computation via instantaneous projection. In addition, we identify the thermodynamic work as the optimal control potential that solves the global variational problem on path space. This perspective unifies non-equilibrium thermodynamics with the geometric framework of optimal transport, interpreting the Jarzynski equality as a consistency condition within the Donsker-Varadhan variational principle, and elucidates the thermodynamic optimality of PA. --- ### A Context Alignment Pre-processor for Enhancing the Coherence of Human-LLM Dialog Authors: Ding Wei Published: 2026-03-17 Categories: cs.AI arXiv: https://arxiv.org/abs/2603.16052v1 Large language models (LLMs) have made remarkable progress in generating fluent text, but they still face a critical challenge of contextual misalignment in long-term and dynamic dialogue. When human users omit premises, simplify references, or shift context abruptly during interactions with LLMs, the models may fail to capture their actual intentions, producing mechanical or off-topic responses that weaken the collaborative potential of dialogue. To address this problem, this paper proposes a computational framework called the Context Alignment Pre-processor (C.A.P.). Rather than operating during generation, C.A.P. functions as a pre-processing module between user input and response generation. The framework includes three core processes: (1) semantic expansion, which extends a user instruction to a broader semantic span including its premises, literal meaning, and implications; (2) time-weighted context retrieval, which prioritizes recent dialogue history through a temporal decay function approximating human conversational focus; and (3) alignment verification and decision branching, which evaluates whether the dialogue remains on track by measuring the semantic similarity between the current prompt and the weighted historical context. When a significant deviation is detected, C.A.P. initiates a structured clarification protocol to help users and the system recalibrate the conversation. This study presents the architecture and theoretical basis of C.A.P., drawing on cognitive science and Common Ground theory in human-computer interaction. We argue that C.A.P. is not only a technical refinement but also a step toward shifting human-computer dialogue from one-way command-execution patterns to two-way, self-correcting, partnership-based collaboration. Finally, we discuss implementation paths, evaluation methods, and implications for the future design of interactive intelligent systems. --- ### The Era of End-to-End Autonomy: Transitioning from Rule-Based Driving to Large Driving Models Authors: Eduardo Nebot, Julie Stephany Berrio Perez Published: 2026-03-17 Categories: cs.RO, cs.CV, eess.IV arXiv: https://arxiv.org/abs/2603.16050v1 Autonomous driving is undergoing a shift from modular rule based pipelines toward end to end (E2E) learning systems. This paper examines this transition by tracing the evolution from classical sense perceive plan control architectures to large driving models (LDMs) capable of mapping raw sensor input directly to driving actions. We analyze recent developments including Tesla's Full Self Driving (FSD) V12 V14, Rivian's Unified Intelligence platform, NVIDIA Cosmos, and emerging commercial robotaxi deployments, focusing on architectural design, deployment strategies, safety considerations and industry implications. A key emerging product category is supervised E2E driving, often referred to as FSD (Supervised) or L2 plus plus, which several manufacturers plan to deploy from 2026 onwards. These systems can perform most of the Dynamic Driving Task (DDT) in complex environments while requiring human supervision, shifting the driver's role to safety oversight. Early operational evidence suggests E2E learning handles the long tail distribution of real world driving scenarios and is becoming a dominant commercial strategy. We also discuss how similar architectural advances may extend beyond autonomous vehicles (AV) to other embodied AI systems, including humanoid robotics. --- ### POaaS: Minimal-Edit Prompt Optimization as a Service to Lift Accuracy and Cut Hallucinations on On-Device sLLMs Authors: Jungwoo Shim, Dae Won Kim, Sun Wook Kim, Soo Young Kim, Myungcheol Lee, Jae-geun Cha, Hyunhwa Choi Published: 2026-03-17 Categories: cs.AI arXiv: https://arxiv.org/abs/2603.16045v1 Small language models (sLLMs) are increasingly deployed on-device, where imperfect user prompts--typos, unclear intent, or missing context--can trigger factual errors and hallucinations. Existing automatic prompt optimization (APO) methods were designed for large cloud LLMs and rely on search that often produces long, structured instructions; when executed under an on-device constraint where the same small model must act as optimizer and solver, these pipelines can waste context and even hurt accuracy. We propose POaaS, a minimal-edit prompt optimization layer that routes each query to lightweight specialists (Cleaner, Paraphraser, Fact-Adder) and merges their outputs under strict drift and length constraints, with a conservative skip policy for well-formed prompts. Under a strict fixed-model setting with Llama-3.2-3B-Instruct and Llama-3.1-8B-Instruct, POaaS improves both task accuracy and factuality while representative APO baselines degrade them, and POaaS recovers up to +7.4% under token deletion and mixup. Overall, per-query conservative optimization is a practical alternative to search-heavy APO for on-device sLLMs. --- ### Enhancing Linguistic Generalization of VLA: Fine-Tuning OpenVLA via Synthetic Instruction Augmentation Authors: Dongik Shin Published: 2026-03-17 Categories: cs.AI arXiv: https://arxiv.org/abs/2603.16044v1 Generalization remains a core challenge in embodied AI, as robots must adapt to diverse environments. While OpenVLA represents the State-of-the-Art (SOTA) in Vision-Language-Action models by leveraging large-scale pre-training, its zero-shot performance can be limited when encountering completely new environments. This paper proposes a parameter-efficient fine-tuning strategy to enhance the linguistic generalization of OpenVLA by synthesizing a general instruction set for the Bridge Dataset V2. The paper leverages a Large Language Model (LLM) to generate a rich variety of semantically equivalent but structurally diverse commands for existing trajectories. In this experiment, Low-Rank Adaptation (LoRA) is implemented to fine-tune OpenVLA on augmented pairs, allowing the model to bridge the gap between complex natural language intent and robotic actions. Results demonstrate that the LoRA-enhanced model's robustness, suggesting that enriching the linguistic space of specialized datasets is crucial for embodied agents. --- ### Collaborative Temporal Feature Generation via Critic-Free Reinforcement Learning for Cross-User Sensor-Based Activity Recognition Authors: Xiaozhou Ye, Feng Jiang, Zihan Wang, Xiulai Wang, Yutao Zhang, Kevin I-Kai Wang Published: 2026-03-17 Categories: cs.LG, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2603.16043v1 Human Activity Recognition using wearable inertial sensors is foundational to healthcare monitoring, fitness analytics, and context-aware computing, yet its deployment is hindered by cross-user variability arising from heterogeneous physiological traits, motor habits, and sensor placements. Existing domain generalization approaches either neglect temporal dependencies in sensor streams or depend on impractical target-domain annotations. We propose a different paradigm: modeling generalizable feature extraction as a collaborative sequential generation process governed by reinforcement learning. Our framework, CTFG (Collaborative Temporal Feature Generation), employs a Transformer-based autoregressive generator that incrementally constructs feature token sequences, each conditioned on prior context and the encoded sensor input. The generator is optimized via Group-Relative Policy Optimization, a critic-free algorithm that evaluates each generated sequence against a cohort of alternatives sampled from the same input, deriving advantages through intra-group normalization rather than learned value estimation. This design eliminates the distribution-dependent bias inherent in critic-based methods and provides self-calibrating optimization signals that remain stable across heterogeneous user distributions. A tri-objective reward comprising class discrimination, cross-user invariance, and temporal fidelity jointly shapes the feature space to separate activities, align user distributions, and preserve fine-grained temporal content. Evaluations on the DSADS and PAMAP2 benchmarks demonstrate state-of-the-art cross-user accuracy (88.53\% and 75.22\%), substantial reduction in inter-task training variance, accelerated convergence, and robust generalization under varying action-space dimensionalities. --- ### Shuffling the Stochastic Mirror Descent via Dual Lipschitz Continuity and Kernel Conditioning Authors: Junwen Qiu, Leilei Mei, Junyu Zhang Published: 2026-03-17 Categories: math.OC, cs.LG, stat.ML arXiv: https://arxiv.org/abs/2603.16042v1 The global Lipschitz smoothness condition underlies most convergence and complexity analyses via two key consequences: the descent lemma and the gradient Lipschitz continuity. How to study the performance of optimization algorithms in the absence of Lipschitz smoothness remains an active area. The relative smoothness framework from Bauschke-Bolte-Teboulle (2017) and Lu-Freund-Nesterov (2018) provides an extended descent lemma, ensuring convergence of Bregman-based proximal gradient methods and their vanilla stochastic counterparts. However, many widely used techniques (e.g., momentum schemes, random reshuffling, and variance reduction) additionally require the Lipschitz-type bound for gradient deviations, leaving their analysis under relative smoothness an open area. To resolve this issue, we introduce the dual kernel conditioning (DKC) regularity condition to regulate the local relative curvature of the kernel functions. Combined with the relative smoothness, DKC provides a dual Lipschitz continuity for gradients: even though the gradient mapping is not Lipschitz in the primal space, it preserves Lipschitz continuity in the dual space induced by a mirror map. We verify that DKC is widely satisfied by popular kernels and is closed under affine composition and conic combination. With these novel tools, we establish the first complexity bounds as well as the iterate convergence of random reshuffling mirror descent for constrained nonconvex relative smooth problems. --- ### Power Analysis for Prediction-Powered Inference Authors: Yiqun T. Chen, Moran Guo, Shengy Li Published: 2026-03-17 Categories: stat.ME, cs.LG arXiv: https://arxiv.org/abs/2603.16041v1 Modern studies increasingly leverage outcomes predicted by machine learning and artificial intelligence (AI/ML) models, and recent work, such as prediction-powered inference (PPI), has developed valid downstream statistical inference procedures. However, classical power and sample size formulas do not readily account for these predictions. In this work, we tackle a simple yet practical question: given a new AI/ML model with high predictive power, how many labeled samples are needed to achieve a desired level of statistical power? We derive closed-form power formulas by characterizing the asymptotic variance of the PPI estimator and applying Wald test inversion to obtain the required labeled sample size. Our results cover widely used settings including two-sample comparisons and risk measures in 2x2 tables. We find that a useful rule of thumb is that the reduction in required labeled samples relative to classical designs scales roughly with the R2 between the predictions and the ground truth. Our analytical formulas are validated using Monte Carlo simulations, and we illustrate the framework in three contemporary biomedical applications spanning single-cell transcriptomics, clinical blood pressure measurement, and dermoscopy imaging. We provide our software as an R package and online calculators at https://github.com/yiqunchen/pppower. --- ### Residual Stream Duality in Modern Transformer Architectures Authors: Yifan Zhang Published: 2026-03-17 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2603.16039v2 Recent work has made clear that the residual pathway is not mere optimization plumbing; it is part of the model's representational machinery. We agree, but argue that the cleanest way to organize this design space is through a two-axis view of the Transformer. A decoder evolves information along two ordered dimensions: sequence position and layer depth. Self-attention already provides adaptive mixing along the sequence axis, whereas the residual stream usually performs fixed addition along the depth axis. If we fix a token position and treat layer index as the ordered variable, then a causal depth-wise residual attention read is exactly the same local operator as causal short sliding-window attention (ShortSWA), except written over depth rather than over sequence. This is the core residual stream duality behind Transformer$^2$. This perspective also clarifies the recent literature. ELC-BERT and DenseFormer already show that learned aggregation over depth can outperform uniform residual accumulation, while Vertical Attention, DeepCrossAttention (DCA), MUDDFormer, and Attention Residuals move further toward explicit attention-based routing over earlier layers. The key point, however, is that operator-level duality does not imply systems-level symmetry. For large-scale autoregressive models, sequence-axis ShortSWA is usually the more hardware-friendly placement because it reuses token-side sliding-window kernels, KV-cache layouts, and chunked execution. If the goal is instead to change the shortcut itself, Deep Delta Learning (DDL) is the cleaner intervention because it modifies the residual operator directly rather than adding a separate cross-layer retrieval path. Our recommendation is therefore simple: use DDL when the shortcut is the object of interest, and use sequence-axis ShortSWA when the goal is local adaptive mixing. --- ### 3D tomography of exchange phase in a Si/SiGe quantum dot device Authors: Dylan Albrecht, Sarah Thompson, N. Tobias Jacobson, Ryan Jock Published: 2026-03-17 Categories: cond-mat.mes-hall, cs.CV, quant-ph arXiv: https://arxiv.org/abs/2603.16025v2 The exchange interaction is a foundational building block for the operation of spin-based quantum processors. Extracting the exchange interaction coefficient $J(\mathbf{V})$, as a function of gate electrode voltages, is important for understanding disorder, faithfully simulating device performance, and operating spin qubits with high fidelity. Typical coherent measurements of exchange in spin qubit devices yield a modulated cosine of an accumulated phase, which in turn is the time integral of exchange. As such, extracting $J(\mathbf{V})$ from experimental data is difficult due to the ambiguity of inverting a cosine, the sensitivity to noise when unwrapping phase, as well as the problem of inverting the integral. As a step toward obtaining $J(\mathbf{V})$, we tackle the first two challenges to reveal the accumulated phase, $φ(\mathbf{V})$. We incorporate techniques from a wide range of fields to robustly extract and model a 3D phase volume for spin qubit devices from a sequence of 2D measurements. In particular, we present a measurement technique to obtain the wrapped phase, as done in phase-shifting digital holography, and utilize the max-flow/min-cut phase unwrapping method (PUMA) to unwrap the phase in 3D voltage space. We show this method is robust to the minimal observed drift in the device, which we confirm by increasing scan resolution. Upon building a model of the extracted phase, we optimize over the model to locate a minimal-gradient $π$ exchange pulse point in voltage space. Our measurement protocol may provide detailed information useful for understanding the origins of device variability governing device yield, enable calibrating device models to specific devices during operation for more sophisticated error attribution, and enable a systematic optimization of qubit control. We anticipate that the methods presented here may be applicable to other qubit platforms. --- ### Speak, Segment, Track, Navigate: An Interactive System for Video-Guided Skull-Base Surgery Authors: Jecia Z. Y. Mao, Francis X. Creighton, Russell H. Taylor, Manish Sahu Published: 2026-03-17 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.16024v2 We introduce a speech-guided embodied agent framework for video-guided skull base surgery that dynamically executes perception and image-guidance tasks in response to surgeon queries. The proposed system integrates natural language interaction with real-time visual perception directly on live intraoperative video streams, thereby enabling surgeons to request computational assistance without disengaging from operative tasks. Unlike conventional image-guided navigation systems that rely on external optical trackers and additional hardware setup, the framework operates purely on intraoperative video. The system begins with interactive segmentation and labeling of the surgical instrument. The segmented instrument is then used as a spatial anchor that is autonomously tracked in the video stream to support downstream workflows, including anatomical segmentation, interactive registration of preoperative 3D models, monocular video-based estimation of the surgical tool pose, and image guidance through real-time anatomical overlays. We evaluate the proposed system in video-guided skull base surgery scenarios and benchmark its tracking performance against a commercially available optical tracking system. Across three experimental trials, the hybrid vision-based method achieved a mean absolute tool-tip position error of 2.32 Plus/Minus 1.10 mm in the camera frame, with inter-frame yaw and pitch propagation discrepancies of 0.18 Plus/Minus 0.25° and 0.21 Plus/Minus 0.30°, respectively. The system completes tool segmentation and anatomy registration within approximately two minutes, substantially reducing setup complexity relative to conventional tracking workflows. These results demonstrate that speech-guided embodied agents can provide accurate spatial guidance while improving workflow integration and enabling rapid deployment of video-guided surgical systems. --- ### Interpretable Context Methodology: Folder Structure as Agentic Architecture Authors: Jake Van Clief, David McDermott Published: 2026-03-17 Categories: cs.AI, cs.HC arXiv: https://arxiv.org/abs/2603.16021v2 Current approaches to AI agent orchestration typically involve building multi-agent frameworks that manage context passing, memory, error handling, and step coordination through code. These frameworks work well for complex, concurrent systems. But for sequential workflows where a human reviews output at each step, they introduce engineering overhead that the problem does not require. This paper presents Model Workspace Protocol (MWP), a method that replaces framework-level orchestration with filesystem structure. Numbered folders represent stages. Plain markdown files carry the prompts and context that tell a single AI agent what role to play at each step. Local scripts handle the mechanical work that does not need AI at all. The result is a system where one agent, reading the right files at the right moment, does the work that would otherwise require a multi-agent framework. This approach applies ideas from Unix pipeline design, modular decomposition, multi-pass compilation, and literate programming to the specific problem of structuring context for AI agents. The protocol is open source under the MIT license. --- ### A Two-Dimensional Framework for AI Agent Design Patterns: Cognitive Function and Execution Topology Authors: Jia Huang, Joey Tianyi Zhou Published: 2026-03-16 Categories: cs.AI, cs.MA, cs.SE arXiv: https://arxiv.org/abs/2605.13850v2 Existing frameworks for LLM-based agent architectures describe systems from a single perspective: industry guides (Anthropic, Google, LangChain) focus on execution topology -- how data flows -- while cognitive science surveys focus on cognitive function -- what the agent does. Neither axis alone disambiguates architecturally distinct systems: the same Orchestrator-Workers topology can implement Plan-and-Execute, Hierarchical Delegation, or Adversarial Verification -- three patterns with fundamentally different failure modes and design trade-offs. We propose a two-dimensional classification that combines (1) a Cognitive Function axis with seven categories (Perception, Memory, Reasoning, Action, Reflection, Collaboration, Governance) and (2) an Execution Topology axis with six structural archetypes (Chain, Route, Parallel, Orchestrate, Loop, Hierarchy). The resulting 7x6 matrix identifies 28 named patterns, 15 with original names. We demonstrate orthogonality through systematic cross-axis analysis, define eight representative patterns in detail, and validate descriptive coverage across four real-world domains (financial lending, legal due diligence, network operations, healthcare triage). Cross-domain analysis yields five empirical laws of pattern selection governing the relationship between environmental constraints (time pressure, action authority, failure cost asymmetry, volume) and architectural choices. The framework provides a principled, framework-neutral, and model-agnostic vocabulary for AI agent architecture design. --- ### Scale-Gest: Scalable Model-Space Synthesis and Runtime Selection for On-Device Gesture Detection Authors: Abdul Basit, Saim Rehman, Muhammad Shafique Published: 2026-03-16 Categories: cs.CV, cs.AI, cs.HC, cs.RO, eess.IV arXiv: https://arxiv.org/abs/2605.12506v1 Realizing on-device ML-based gesture detection under tight real-time performance, energy and memory constraints is challenging, especially when considering mobile devices with varying battery-power levels. Existing EdgeAI deployments typically rely on a single fixed detector, limiting optimization opportunities. We present Scale-Gest, a novel run-time adaptive gesture detection framework that expands the detector space into a dense family of tiny-YOLO architectures. We introduce multiple novel device-calibrated ACE (Accuracy-Complexity-Energy) profiles by analyzing different model-resolution-stride operating points. A lightweight run-time controller selects an appropriate ACE mode under user-defined and battery constraints, while a motion-aware hand-gesture-tracking ROI gate crops the input for reduced complexity detection. To evaluate performance of our system in real-world car driving scenarios, we introduce a temporally-annotated Driver Simulated Gesture (DSG-18) dataset. Scale-Gest maintains event-level F1 while significantly reducing energy and latency compared to single-detector approaches. On a battery-powered laptop running gesture streams, our ACE controller reduces per-frame energy by 4x (from 6.9 mJ to 1.6 mJ) while maintaining high gesture-detection performance (event-level F1 = 0.8-0.9) and low mean latency (6 ms). --- ### AI to Learn 2.0: A Deliverable-Oriented Governance Framework and Maturity Rubric for Opaque AI in Learning-Intensive Domains Authors: Seine A. Shintani Published: 2026-03-16 Categories: cs.AI, cs.CY arXiv: https://arxiv.org/abs/2604.19751v1 Generative AI is entering research, education, and professional work faster than current governance frameworks can specify how AI-assisted outputs should be judged in learning-intensive settings. The central problem is proxy failure: a polished artifact can be useful while no longer serving as credible evidence of the human understanding, judgment, or transfer ability that the work is supposed to cultivate or certify. This paper proposes AI to Learn 2.0, a deliverable-oriented governance framework for AI-assisted work. Rather than claiming element-wise novelty, it reorganizes adjacent ideas around the final deliverable package, distinguishes artifact residual from capability residual, and operationalizes the result through a five-part package, a seven-dimension maturity rubric, gate thresholds on critical dimensions, and a companion capability-evidence ladder. AI to Learn 2.0 allows opaque AI during exploration, drafting, hypothesis generation, and workflow design, but requires that the released deliverable be usable, auditable, transferable, and justifiable without the original large language model or cloud API. In learning-intensive contexts, it additionally requires context-appropriate human-attributable evidence of explanation or transfer. Worked scoring across contrastive cases, including coursework substitution, a symbolic-regression governance contrast, teacher-audited national-exam practice forms, and a self-hosted lecture-to-quiz pipeline with deterministic quality control, shows how the framework separates polished substitution workflows from bounded, auditable, and handoff-ready AI-assisted workflows. AI to Learn 2.0 is proposed as a governance instrument for structured third-party review where capability preservation, accountability, and validity boundaries matter. --- ### Training for Compositional Sensitivity Reduces Dense Retrieval Generalization Authors: Radoslav Ralev, Aditeya Baral, Iliya Zhechev, Jen Agarwal, Srijith Rajamohan Published: 2026-03-16 Categories: cs.IR, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2604.16351v1 Dense retrieval compresses texts into single embeddings ranked by cosine similarity. While efficient for recall, this interface is brittle for identity-level matching: minimal compositional edits (negation, role swaps) flip meaning yet retain high similarity. Motivated by geometric results for unit-sphere cosine spaces (Kang et al., 2025), we test this retrieval-composition tension in text-only retrieval. Across four dual-encoder backbones, adding structure-targeted negatives consistently reduces zero-shot NanoBEIR retrieval (8-9% mean nDCG@10 drop on small backbones; up to 40% on medium ones), while only partially improving pooled-space separation. Treating pooled cosine as a recall interface, we then benchmark verifiers scoring token--token cosine maps. MaxSim (late interaction) excels at reranking but fails to reject structural near-misses, whereas a small Transformer over similarity maps reliably separates near-misses under end-to-end training. --- ### LiteSemRAG: Lightweight LLM-Free Semantic-Aware Graph Retrieval for Robust RAG Authors: Xiao Yue, Guangzhi Qu, Lige Gan Published: 2026-03-16 Categories: cs.IR, cs.AI arXiv: https://arxiv.org/abs/2604.16350v1 Graph-based Retrieval-Augmented Generation (RAG) has shown great potential for improving multi-level reasoning and structured evidence aggregation. However, existing graph-based RAG frameworks heavily rely on exploiting large language models (LLMs) during indexing and querying, leading to high token consumption, computational cost and latency overhead. In this paper, we propose LiteSemRAG, a lightweight, fully LLM-free, semantic-aware graph retrieval framework. LiteSemRAG constructs a heterogeneous semantic graph by exploiting contextual token-level embeddings, explicitly separating surface lexical representations from context-dependent semantic meanings. To robustly model polysemy, we introduce a dynamic semantic node construction mechanism with chunk-level context aggregation and adaptive anomaly handling. At query stage, LiteSemRAG performs a two-step semantic-aware retrieval process that integrates co-occurrence graph weighting with an isolated semantic recovery mechanism, enabling balanced structural reasoning and semantic coverage. We evaluate LiteSemRAG on three benchmark datasets and experimental results show that LiteSemRAG achieves the best mean reciprocal rank (MRR@10) across all datasets and competitive or superior recall rate (Recall@10) compared to state-of-the-art LLM-based graph RAG systems. Meanwhile, LiteSemRAG consumes zero LLM tokens and achieves substantial efficiency improvements in both indexing and querying due to the elimination of LLM usage. These results demonstrate the effectiveness of LiteSemRAG, indicating that a strong semantic-aware graph retrieval framework can be achieved without relying on LLM-based approaches. --- ### Benchmarking Real-Time Question Answering via Executable Code Workflows Authors: Wenjie Zhou, Yuan Gao, Xin Zhou, Hao Fu, Zhongjian Miao, Wei Chen, Bo Chen, Xiaobing Zhao Published: 2026-03-16 Categories: cs.IR, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2604.16349v2 Retrieving real-time information is a fundamental capability for search-integrated agents in real-world applications. However, existing benchmarks are predominantly static and therefore fail to capture the temporal dynamics of information and the continuously evolving nature of real-world knowledge. To address this limitation, we propose RT-QA, a dynamic evaluation framework that leverages executable code workflows to retrieve up-to-date answers at evaluation time. Specifically, we construct an agent-driven pipeline that autonomously generates code for web crawling and DOM-based answer extraction to produce real-time ground truth. To ensure robust evaluation over time, the pipeline further incorporates a self-repair mechanism to adapt to changes in web page structures. RT-QA spans 12 domains (e.g., Finance, Sports) with 320 Chinese questions categorized into three difficulty levels. Extensive evaluations of state-of-the-art models (e.g., GPT-5.2, GLM-4.7) reveal significant limitations in real-time adaptability: even the best models achieve only 46% accuracy. Our analysis highlights two primary failure modes: (1) Lazy Retrieval, where agents rely on search snippets instead of deeply scanning specific websites for information (20% of failures); and (2) Temporal Confusion, a cognitive error where agents retrieve a historical date (e.g., an event in 2024) and fail to re-anchor to the current time (2026) for subsequent reasoning. These findings suggest that future agents require not just better retrieval strategies, but robust temporal state management. ---