# Research Papers — 2026-04 · bundle page 133 of 233 > Full-text markdown of papers #6601–6650 of 11612 published in 2026-04, newest first. AI-readable mirrors of the full 11612-paper 2026-04 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-04/page/133 [← Previous bundle](https://aipod.dev/papers/archive/2026-04/page/132.md) [Next bundle →](https://aipod.dev/papers/archive/2026-04/page/134.md) ### How Transformers Learn to Plan via Multi-Token Prediction Authors: Jianhao Huang, Zhanpeng Zhou, Renqiu Xia, Baharan Mirzasoleiman, Weijie Su, Wei Huang Published: 2026-04-13 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2604.11912v2 While next-token prediction (NTP) has been the standard objective for training language models, it often struggles to capture global structure in reasoning tasks. Multi-token prediction (MTP) has recently emerged as a promising alternative, yet its underlying mechanisms remain poorly understood. In this paper, we study how MTP facilitates reasoning, with a focus on planning. Empirically, we show that MTP consistently outperforms NTP on both synthetic graph path-finding tasks and more realistic reasoning benchmarks, such as Countdown and boolean satisfiability problems. Theoretically, we analyze a simplified two-layer Transformer on a star graph task. We prove that MTP induces a two-stage reverse reasoning process: the model first attends to the end node and then reconstructs the path by tracing intermediate nodes backward. This behavior arises from a gradient decoupling property of MTP, which provides a cleaner training signal compared to NTP. Ultimately, our results highlight how multi-token objectives inherently bias optimization toward robust and interpretable reasoning circuits. --- ### Thermodynamic Liquid Manifold Networks: Physics-Bounded Deep Learning for Solar Forecasting in Autonomous Off-Grid Microgrids Authors: Mohammed Ezzaldin Babiker Abdullah Published: 2026-04-13 Categories: cs.LG, cs.AI, eess.SY arXiv: https://arxiv.org/abs/2604.11909v2 The stable operation of autonomous off-grid photovoltaic systems requires solar forecasting algorithms that respect atmospheric thermodynamics. Contemporary deep learning models consistently exhibit critical anomalies, primarily severe temporal phase lags during cloud transients and physically impossible nocturnal power generation. To resolve this divergence between data-driven modeling and deterministic celestial mechanics, this research introduces the Thermodynamic Liquid Manifold Network. The methodology projects 22 meteorological and geometric variables into a Koopman-linearized Riemannian manifold to systematically map complex climatic dynamics. The architecture integrates a Spectral Calibration unit and a multiplicative Thermodynamic Alpha-Gate. This system synthesizes real-time atmospheric opacity with theoretical clear-sky boundary models, structurally enforcing strict celestial geometry compliance. This completely neutralizes phantom nocturnal generation while maintaining zero-lag synchronization during rapid weather shifts. Validated against a rigorous five-year testing horizon in a severe semi-arid climate, the framework achieves an RMSE of 18.31 Wh/m2 and a Pearson correlation of 0.988. The model strictly maintains a zero-magnitude nocturnal error across all 1826 testing days and exhibits a sub-30-minute phase response during high-frequency optical transients. Comprising exactly 63,458 trainable parameters, this ultra-lightweight design establishes a robust, thermodynamically consistent standard for edge-deployable microgrid controllers. --- ### Subcritical Signal Propagation at Initialization in Normalization-Free Transformers Authors: Sergey Alekseev Published: 2026-04-13 Categories: cs.LG, stat.ML arXiv: https://arxiv.org/abs/2604.11890v3 We study signal propagation at initialization in transformers through the averaged partial Jacobian norm (APJN), a measure of gradient amplification across layers. We extend APJN analysis to transformers with bidirectional attention and permutation-symmetric input token configurations by deriving recurrence relations for activation statistics and APJNs across layers. Our theory predicts how attention modifies the asymptotic behavior of the APJN at large depth and matches APJNs measured in deep vision transformers. The criticality picture known from residual networks carries over to transformers: the pre-LayerNorm architecture exhibits power-law APJN growth, whereas transformers with LayerNorm replaced by elementwise $\tanh$-like nonlinearities have stretched-exponential APJN growth, indicating that the latter are subcritical. Applied to Dynamic Tanh (DyT) and Dynamic erf (Derf) transformers, the theory explains why these architectures can be more sensitive to initialization and optimization choices and require careful tuning for stable training. --- ### MedConcept: Unsupervised Concept Discovery for Interpretability in Medical VLMs Authors: Md Rakibul Haque, KM Arefeen Sultan, Tushar Kataria, Shireen Elhabian Published: 2026-04-13 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.11868v1 While medical Vision-Language models (VLMs) achieve strong performance on tasks such as tumor or organ segmentation and diagnosis prediction, their opaque latent representations limit clinical trust and the ability to explain predictions. Interpretability of these multimodal representations are therefore essential for the trustworthy clinical deployment of pretrained medical VLMs. However, current interpretability methods, such as gradient- or attention-based visualizations, are often limited to specific tasks such as classification. Moreover, they do not provide concept-level explanations derived from shared pretrained representations that can be reused across downstream tasks. We introduce MedConcept, a framework that uncovers latent medical concepts in a fully unsupervised manner and grounds them in clinically verifiable textual semantics. MedConcept identifies sparse neuron-level concept activations from pretrained VLM representations and translates them into pseudo-report-style summaries, enabling physician-level inspection of internal model reasoning. To address the lack of quantitative evaluation in concept-based interpretability, we introduce a quantitative semantic verification protocol that leverages an independent pretrained medical LLM as a frozen external evaluator to assess concept alignment with radiology reports. We define three concept scores, Aligned, Unaligned, and Uncertain, to quantify semantic support, contradiction, or ambiguity relative to radiology reports and use them exclusively for post hoc evaluation. These scores provide a quantitative baseline for assessing interpretability in medical VLMs. All codes, prompt and data to be released on acceptance. Ke --- ### Disposition Distillation at Small Scale: A Three-Arc Negative Result Authors: Hari Sadasivan Published: 2026-04-13 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2604.11867v1 We set out to train behavioral dispositions (self-verification, uncertainty acknowledgment, feedback integration) into small language models (0.6B to 2.3B effective parameters) through a four-stage all-MIT distillation pipeline, with follow-on experiments on inference-time attention-head interventions and a frozen-base confidence-gated sidecar. An internal draft reported +33.9-point MCAS and +15.3-point HumanEval gains on a Qwen3-0.6B student; a second-pass sanity check falsified both numbers before publication. The HumanEval delta was a truncation artifact (n_predict=512) that inverted to -8.0 points at n_predict=1024; the MCAS gain disappeared under apples-to-apples scoring. That falsification triggered three subsequent arcs. Across (1) SFT/DPO LoRA on three model families and two domains, (2) inference-time attention-head tempering on o_proj, and (3) a training-free frozen-base sidecar reading the final-token hidden state h_last, we find no operator that moves judge-measured disposition without damaging content or collapsing into stylistic mimicry. The failure is consistent across five models (Qwen3-0.6B, Qwen3-1.7B, Qwen3.5-0.8B, Gemma 4 E2B, and SmolLM2-1.7B-Instruct). A within-distribution cross-validation pass (AUC=0.683) collapsed to chance on fresh prompts (AUC=0.516). We contribute a three-arc negative result with mechanism, a two-failure-mode taxonomy for linear h_last probes, and an honest falsification pipeline that converts the class of false positives we ourselves produced into publishable negatives. As an independent finding, Gemma 4 E2B exhibits near-complete confidence-correctness decoupling on the Chef domain (assertion asymmetry -0.009; the model asserts at 91% regardless of correctness). --- ### Obtaining Partition Crossover masks using Statistical Linkage Learning for solving noised optimization problems with hidden variable dependency structure Authors: M. W. Przewozniczek, B. Frej, M. M. Komarnicki, M. Prusik, R. Tinós Published: 2026-04-13 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2604.11862v1 In optimization problems, some variable subsets may have a joint non-linear or non-monotonical influence on the function value. Therefore, knowledge of variable dependencies may be crucial for effective optimization, and many state-of-the-art optimizers leverage it to improve performance. However, some real-world problem instances may be the subject of noise of various origins. In such a case, variable dependencies relevant to optimization may be hard or impossible to tell using dependency checks sufficient for problems without noise, making highly effective operators, e.g., Partition Crossover (PX), useless. Therefore, we use Statistical Linkage Learning (SLL) to decompose problems with noise and propose a new SLL-dedicated mask construction algorithm. We prove that if the quality of the SLL-based decomposition is sufficiently high, the proposed clustering algorithm yields masks equivalent to PX masks for the noise-free instances. The experiments show that the optimizer using the proposed mechanisms remains equally effective despite the noise level and outperforms state-of-the-art optimizers for the problems with high noise. --- ### MVAdapt: Zero-Shot Multi-Vehicle Adaptation for End-to-End Autonomous Driving Authors: Haesung Oh, Jaeheung Park Published: 2026-04-13 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2604.11854v1 End-to-End (E2E) autonomous driving models are usually trained and evaluated with a fixed ego-vehicle, even though their driving policy is implicitly tied to vehicle dynamics. When such a model is deployed on a vehicle with different size, mass, or drivetrain characteristics, its performance can degrade substantially; we refer to this problem as the vehicle-domain gap. To address it, we propose MVAdapt, a physics-conditioned adaptation framework for multi-vehicle E2E driving. MVAdapt combines a frozen TransFuser++ scene encoder with a lightweight physics encoder and a cross-attention module that conditions scene features on vehicle properties before waypoint decoding. In the CARLA Leaderboard 1.0 benchmark, MVAdapt improves over naive transfer and multi-embodiment adaptation baselines on both in-distribution and unseen vehicles. We further show two complementary behaviors: strong zero-shot transfer on many unseen vehicles, and data-efficient few-shot calibration for severe physical outliers. These results suggest that explicitly conditioning E2E driving policies on vehicle physics is an effective step toward more transferable autonomous driving models. All codes are available at https://github.com/hae-sung-oh/MVAdapt --- ### Limitations of Sequence-Based Protein Representations for Parkinson's Disease Classification: A Leakage-Free Benchmark Authors: César Jesús Núñez-Prado, Grigori Sidorov, Liliana Chanona-Hernández Published: 2026-04-13 Categories: q-bio.QM, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2604.11852v2 The identification of reliable molecular biomarkers for Parkinson's disease remains challenging due to its multifactorial nature. Although protein sequences constitute a fundamental and widely available source of biological information, their standalone discriminative capacity for complex disease classification remains unclear. In this work, we present a controlled and leakage-free evaluation of multiple representations derived exclusively from protein primary sequences, including amino acid composition, k-mers, physicochemical descriptors, hybrid representations, and embeddings from protein language models, all assessed under a nested stratified cross-validation framework to ensure unbiased performance estimation. The best-performing configuration (ProtBERT + MLP) achieves an F1-score of 0.704 +/- 0.028 and ROC-AUC of 0.748 +/- 0.047, indicating only moderate discriminative performance. Classical representations such as k-mers reach comparable F1 values (up to approximately 0.667), but exhibit highly imbalanced behavior, with recall close to 0.98 and precision around 0.50, reflecting a strong bias toward positive predictions. Across representations, performance differences remain within a narrow range (F1 between 0.60 and 0.70), while unsupervised analyses reveal no intrinsic structure aligned with class labels, and statistical testing (Friedman test, p = 0.1749) does not indicate significant differences across models. These results demonstrate substantial overlap between classes and indicate that primary sequence information alone provides limited discriminative power for Parkinson's disease classification. This work establishes a reproducible baseline and provides empirical evidence that more informative biological features, such as structural, functional, or interaction-based descriptors, are required for robust disease modeling. --- ### Who Handles Orientation? Investigating Invariance in Feature Matching Authors: David Nordström, Johan Edstedt, Fredrik Kahl, Georg Bökman Published: 2026-04-13 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.11809v1 Finding matching keypoints between images is a core problem in 3D computer vision. However, modern matchers struggle with large in-plane rotations. A straightforward mitigation is to learn rotation invariance via data augmentation. However, it remains unclear at which stage rotation invariance should be incorporated. In this paper, we study this in the context of a modern sparse matching pipeline. We perform extensive experiments by training on a large collection of 3D vision datasets and evaluating on popular image matching benchmarks. Surprisingly, we find that incorporating rotation invariance already in the descriptor yields similar performance to handling it in the matcher. However, rotation invariance is achieved earlier in the matcher when it is learned in the descriptor, allowing for a faster rotation-invariant matcher. Further, we find that enforcing rotation invariance does not hurt upright performance when trained at scale. Finally, we study the emergence of rotation invariance through scale and find that increasing the training data size substantially improves generalization to rotated images. We release two matchers robust to in-plane rotations that achieve state-of-the-art performance on e.g. multi-modal (WxBS), extreme (HardMatch), and satellite image matching (SatAst). Code is available at https://github.com/davnords/loma. --- ### Pair2Scene: Learning Local Object Relations for Procedural Scene Generation Authors: Xingjian Ran, Shujie Zhang, Weipeng Zhong, Li Luo, Bo Dai Published: 2026-04-13 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.11808v2 Generating high-fidelity 3D indoor scenes remains a significant challenge due to data scarcity and the complexity of modeling intricate spatial relations. Current methods often struggle to scale beyond training distribution to dense scenes or rely on LLMs/VLMs that lack the ability for precise spatial reasoning. Building on top of the observation that object placement relies mainly on local dependencies instead of information-redundant global distributions, in this paper, we propose Pair2Scene, a novel procedural generation framework that integrates learned local rules with scene hierarchies and physics-based algorithms. These rules mainly capture two types of inter-object relations, namely support relations that follow physical hierarchies, and functional relations that reflect semantic links. We model these rules through a network, which estimates spatial position distributions of dependent objects conditioned on position and geometry of the anchor ones. Accordingly, we curate a dataset 3D-Pairs from existing scene data to train the model. During inference, our framework can generate scenes by recursively applying our model within a hierarchical structure, leveraging collision-aware rejection sampling to align local rules into coherent global layouts. Extensive experiments demonstrate that our framework outperforms existing methods in generating complex environments that go beyond training data while maintaining physical and semantic plausibility. --- ### Physics-Informed State Space Models for Reliable Solar Irradiance Forecasting in Off-Grid Systems Authors: Mohammed Ezzaldin Babiker Abdullah Published: 2026-04-13 Categories: cs.LG, cs.AI, eess.SY arXiv: https://arxiv.org/abs/2604.11807v3 The stable operation of off-grid photovoltaic systems requires accurate, computationally efficient solar forecasting. Contemporary deep learning models often suffer from massive computational overhead and physical blindness, generating impossible predictions. This paper introduces the Physics-Informed State Space Model (PISSM) to bridge the gap between efficiency and physical accuracy for edge-deployed microcontrollers. PISSM utilizes a dynamic Hankel matrix embedding to filter stochastic sensor noise by transforming raw meteorological sequences into a robust state space. A Linear State Space Model replaces heavy attention mechanisms, efficiently modeling temporal dependencies for parallel processing. Crucially, a novel Physics-Informed Gating mechanism leverages the Solar Zenith Angle and Clearness Index to structurally bound outputs, ensuring predictions strictly obey diurnal cycles and preventing nocturnal errors. Validated on a multi-year dataset for Omdurman, Sudan, PISSM achieves superior accuracy with fewer than 40,000 parameters, establishing an ultra-lightweight benchmark for real-time off-grid control. --- ### Detecting Safety Violations Across Many Agent Traces Authors: Adam Stein, Davis Brown, Hamed Hassani, Mayur Naik, Eric Wong Published: 2026-04-13 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2604.11806v1 To identify safety violations, auditors often search over large sets of agent traces. This search is difficult because failures are often rare, complex, and sometimes even adversarially hidden and only detectable when multiple traces are analyzed together. These challenges arise in diverse settings such as misuse campaigns, covert sabotage, reward hacking, and prompt injection. Existing approaches struggle here for several reasons. Per-trace judges miss failures that only become visible across traces, naive agentic auditing does not scale to large trace collections, and fixed monitors are brittle to unanticipated behaviors. We introduce Meerkat, which combines clustering with agentic search to uncover violations specified in natural language. Through structured search and adaptive investigation of promising regions, Meerkat finds sparse failures without relying on seed scenarios, fixed workflows, or exhaustive enumeration. Across misuse, misalignment, and task gaming settings, Meerkat significantly improves detection of safety violations over baseline monitors, discovers widespread developer cheating on a top agent benchmark, and finds nearly 4x more examples of reward hacking on CyBench than previous audits. --- ### Solving Physics Olympiad via Reinforcement Learning on Physics Simulators Authors: Mihir Prabhudesai, Aryan Satpathy, Yangmin Li, Zheyang Qin, Nikash Bhardwaj, Amir Zadeh, Chuan Li, Katerina Fragkiadaki, Deepak Pathak Published: 2026-04-13 Categories: cs.LG, cs.AI, cs.CV, cs.RO arXiv: https://arxiv.org/abs/2604.11805v1 We have witnessed remarkable advances in LLM reasoning capabilities with the advent of DeepSeek-R1. However, much of this progress has been fueled by the abundance of internet question-answer (QA) pairs, a major bottleneck going forward, since such data is limited in scale and concentrated mainly in domains like mathematics. In contrast, other sciences such as physics lack large-scale QA datasets to effectively train reasoning-capable models. In this work, we show that physics simulators can serve as a powerful alternative source of supervision for training LLMs for physical reasoning. We generate random scenes in physics engines, create synthetic question-answer pairs from simulated interactions, and train LLMs using reinforcement learning on this synthetic data. Our models exhibit zero-shot sim-to-real transfer to real-world physics benchmarks: for example, training solely on synthetic simulated data improves performance on IPhO (International Physics Olympiad) problems by 5-10 percentage points across model sizes. These results demonstrate that physics simulators can act as scalable data generators, enabling LLMs to acquire deep physical reasoning skills beyond the limitations of internet-scale QA data. Code available at: https://sim2reason.github.io/. --- ### OmniShow: Unifying Multimodal Conditions for Human-Object Interaction Video Generation Authors: Donghao Zhou, Guisheng Liu, Hao Yang, Jiatong Li, Jingyu Lin, Xiaohu Huang, Yichen Liu, Xin Gao, Cunjian Chen, Shilei Wen, Chi-Wing Fu, Pheng-Ann Heng Published: 2026-04-13 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.11804v2 In this work, we study Human-Object Interaction Video Generation (HOIVG), which aims to synthesize high-quality human-object interaction videos conditioned on text, reference images, audio, and pose. This task holds significant practical value for automating content creation in real-world applications, such as e-commerce demonstrations, short video production, and interactive entertainment. However, existing approaches fail to accommodate all these requisite conditions. We present OmniShow, an end-to-end framework tailored for this practical yet challenging task, capable of harmonizing multimodal conditions and delivering industry-grade performance. To overcome the trade-off between controllability and quality, we introduce Unified Channel-wise Conditioning for efficient image and pose injection, and Gated Local-Context Attention to ensure precise audio-visual synchronization. To effectively address data scarcity, we develop a Decoupled-Then-Joint Training strategy that leverages a multi-stage training process with model merging to efficiently harness heterogeneous sub-task datasets. Furthermore, to fill the evaluation gap in this field, we establish HOIVG-Bench, a dedicated and comprehensive benchmark for HOIVG. Extensive experiments demonstrate that OmniShow achieves overall state-of-the-art performance across various multimodal conditioning settings, setting a solid standard for the emerging HOIVG task. --- ### Saar-Voice: A Multi-Speaker Saarbrücken Dialect Speech Corpus Authors: Lena S. Oberkircher, Jesujoba O. Alabi, Dietrich Klakow, Jürgen Trouvain Published: 2026-04-13 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.11803v1 Natural language processing (NLP) and speech technologies have made significant progress in recent years; however, they remain largely focused on standardized language varieties. Dialects, despite their cultural significance and widespread use, are underrepresented in linguistic resources and computational models, resulting in performance disparities. To address this gap, we introduce Saar-Voice, a six-hour speech corpus for the Saarbrücken dialect of German. The dataset was created by first collecting text through digitized books and locally sourced materials. A subset of this text was recorded by nine speakers, and we conducted analyses on both the textual and speech components to assess the dataset's characteristics and quality. We discuss methodological challenges related to orthographic and speaker variation, and explore grapheme-to-phoneme (G2P) conversion. The resulting corpus provides aligned textual and audio representations. This serves as a foundation for future research on dialect-aware text-to-speech (TTS), particularly in low-resource scenarios, including zero-shot and few-shot model adaptation. --- ### Psychological Concept Neurons: Can Neural Control Bias Probing and Shift Generation in LLMs? Authors: Yuto Harada, Hiro Taiyo Hamada Published: 2026-04-13 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.11802v1 Using psychological constructs such as the Big Five, large language models (LLMs) can imitate specific personality profiles and predict a user's personality. While LLMs can exhibit behaviors consistent with these constructs, it remains unclear where and how they are represented inside the model and how they relate to behavioral outputs. To address this gap, we focus on questionnaire-operationalized Big Five concepts, analyze the formation and localization of their internal representations, and use interventions to examine how these representations relate to behavioral outputs. In our experiment, we first use probing to examine where Big Five information emerges across model depth. We then identify neurons that respond selectively to each Big Five concept and test whether enhancing or suppressing their activations can bias latent representations and label generation in intended directions. We find that Big Five information becomes rapidly decodable in early layers and remains detectable through the final layers, while concept-selective neurons are most prevalent in mid layers and exhibit limited overlap across domains. Interventions on these neurons consistently shift probe readouts toward targeted concepts, with targeted success rates exceeding 0.8 for some concepts, indicating that the model's internal separation of Big Five personality traits can be causally steered. At the label-generation level, the same interventions often bias generated label distributions in the intended directions, but the effects are weaker, more concept-dependent, and often accompanied by cross-trait spillover, indicating that comparable control over generated labels is difficult even with interventions on a large fraction of concept-selective neurons. Overall, our findings reveal a gap between representational control and behavioral control in LLMs. --- ### CLSGen: A Dual-Head Fine-Tuning Framework for Joint Probabilistic Classification and Verbalized Explanation Authors: WonJin Yoon, Kangyu Zhu, Ian Bulovic, Autumn Sehy, Yanjun Gao, Dmitriy Dligach, Majid Afshar, Timothy A. Miller Published: 2026-04-13 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.11801v1 With the recent progress of Large Language Models (LLMs), there is a growing interest in applying these models to solve complex and challenging problems. Modern LLMs, capable of processing long contexts and generating verbalized explanations, offer significant potential in addressing real-world applications. However, a critical hurdle in deploying LLMs for practical decision-making is their inability to provide reliable, quantitative probabilities. While task-specific fine-tuning of LLMs using traditional discriminative objectives (similar to encoder-only models) can yield probability estimates, this often leads to catastrophic forgetting and linguistic collapse. Consequently, the model loses its ability to generate explanations, severely undermining its interpretability and usability. To address this challenge, we propose CLSGen, a novel LLM fine-tuning framework designed for binary classification tasks. The CLSGen framework encompasses a new model architecture, training methodology, and data construction strategy to enable robust probability estimation without sacrificing the model's inherent explanation-generation capabilities. Experimental results across multiple benchmark datasets demonstrate that models fine-tuned with CLSGen outperform existing baselines in classification metrics (AUROC and F1-score). Regarding explanation, the results showed strong alignment between predicted labels and generated justifications, as well as high readability. --- ### Budget-Aware Uncertainty for Radiotherapy Segmentation QA Using nnU-Net Authors: Ricardo Coimbra Brioso, Lorenzo Mondo, Damiano Dei, Nicola Lambri, Pietro Mancosu, Marta Scorsetti, Daniele Loiacono Published: 2026-04-13 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2604.11798v1 Accurate delineation of the Clinical Target Volume (CTV) is essential for radiotherapy planning, yet remains time-consuming and difficult to assess, especially for complex treatments such as Total Marrow and Lymph Node Irradiation (TMLI). While deep learning-based auto-segmentation can reduce workload, safe clinical deployment requires reliable cues indicating where models may be wrong. In this work, we propose a budget-aware uncertainty-driven quality assurance (QA) framework built on nnU-Net, combining uncertainty quantification and post-hoc calibration to produce voxel-wise uncertainty maps (based on predictive entropy) that can guide targeted manual review. We compare temperature scaling (TS), deep ensembles (DE), checkpoint ensembles (CE), and test-time augmentation (TTA), evaluated both individually and in combination on TMLI as a representative use case. Reliability is assessed through ROI-masked calibration metrics and uncertainty--error alignment under realistic revision constraints, summarized as AUC over the top 0-5% most uncertain voxels. Across configurations, segmentation accuracy remains stable, whereas TS substantially improves calibration. Uncertainty-error alignment improves most with calibrated checkpoint-based inference, leading to uncertainty maps that highlight more consistently regions requiring manual edits. Overall, integrating calibration with efficient ensembling seems a promising strategy to implement a budget-aware QA workflow for radiotherapy segmentation. --- ### SyncFix: Fixing 3D Reconstructions via Multi-View Synchronization Authors: Deming Li, Abhay Yadav, Cheng Peng, Rama Chellappa, Anand Bhattad Published: 2026-04-13 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.11797v2 We present SyncFix, a framework that enforces cross-view consistency during the diffusion-based refinement of reconstructed scenes. SyncFix formulates refinement as a joint latent bridge matching problem, synchronizing distorted and clean representations across multiple views to fix the semantic and geometric inconsistencies. This means SyncFix learns a joint conditional over multiple views to enforce consistency throughout the denoising trajectory. Our training is done only on image pairs, but it generalizes naturally to an arbitrary number of views during inference. Moreover, reconstruction quality improves with additional views, with diminishing returns at higher view counts. Qualitative and quantitative results demonstrate that SyncFix consistently generates high-quality reconstructions and surpasses current state-of-the-art baselines, even in the absence of clean reference images. SyncFix achieves even higher fidelity when sparse references are available. --- ### C-ReD: A Comprehensive Chinese Benchmark for AI-Generated Text Detection Derived from Real-World Prompts Authors: Chenxi Qing, Junxi Wu, Zheng Liu, Yixiang Qiu, Hongyao Yu, Bin Chen, Hao Wu, Shu-Tao Xia Published: 2026-04-13 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.11796v2 Recently, large language models (LLMs) are capable of generating highly fluent textual content. While they offer significant convenience to humans, they also introduce various risks, like phishing and academic dishonesty. Numerous research efforts have been dedicated to developing algorithms for detecting AI-generated text and constructing relevant datasets. However, in the domain of Chinese corpora, challenges remain, including limited model diversity and data homogeneity. To address these issues, we propose C-ReD: a comprehensive Chinese Real-prompt AI-generated Detection benchmark. Experiments demonstrate that C-ReD not only enables reliable in-domain detection but also supports strong generalization to unseen LLMs and external Chinese datasets-addressing critical gaps in model diversity, domain coverage, and prompt realism that have limited prior Chinese detection benchmarks. We release our resources at https://github.com/HeraldofLight/C-ReD. --- ### LottieGPT: Tokenizing Vector Animation for Autoregressive Generation Authors: Junhao Chen, Kejun Gao, Yuehan Cui, Mingze Sun, Mingjin Chen, Shaohui Wang, Xiaoxiao Long, Fei Ma, Qi Tian, Ruqi Huang, Hao Zhao Published: 2026-04-13 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.11792v1 Despite rapid progress in video generation, existing models are incapable of producing vector animation, a dominant and highly expressive form of multimedia on the Internet. Vector animations offer resolution-independence, compactness, semantic structure, and editable parametric motion representations, yet current generative models operate exclusively in raster space and thus cannot synthesize them. Meanwhile, recent advances in large multimodal models demonstrate strong capabilities in generating structured data such as slides, 3D meshes, LEGO sequences, and indoor layouts, suggesting that native vector animation generation may be achievable. In this work, we present the first framework for tokenizing and autoregressively generating vector animations. We adopt Lottie, a widely deployed JSON-based animation standard, and design a tailored Lottie Tokenizer that encodes layered geometric primitives, transforms, and keyframe-based motion into a compact and semantically aligned token sequence. To support large-scale training, we also construct LottieAnimation-660K, the largest and most diverse vector animation dataset to date, consisting of 660k real-world Lottie animation and 15M static Lottie image files curated from broad Internet sources. Building upon these components, we finetune Qwen-VL to create LottieGPT, a native multimodal model capable of generating coherent, editable vector animations directly from natural language or visual prompts. Experiments show that our tokenizer dramatically reduces sequence length while preserving structural fidelity, enabling effective autoregressive learning of dynamic vector content. LottieGPT exhibits strong generalization across diverse animation styles and outperforms previous state-of-the-art models on SVG generation (a special case of single-frame vector animation). --- ### A Mechanistic Analysis of Looped Reasoning Language Models Authors: Hugh Blayney, Álvaro Arroyo, Johan Obando-Ceron, Pablo Samuel Castro, Aaron Courville, Michael M. Bronstein, Xiaowen Dong Published: 2026-04-13 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2604.11791v1 Reasoning has become a central capability in large language models. Recent research has shown that reasoning performance can be improved by looping an LLM's layers in the latent dimension, resulting in looped reasoning language models. Despite promising results, few works have investigated how their internal dynamics differ from those of standard feedforward models. In this paper, we conduct a mechanistic analysis of the latent states in looped language models, focusing in particular on how the stages of inference observed in feedforward models compare to those observed in looped ones. To this end, we analyze cyclic recurrence and show that for many of the studied models each layer in the cycle converges to a distinct fixed point; consequently, the recurrent block follows a consistent cyclic trajectory in the latent space. We provide evidence that as these fixed points are reached, attention-head behavior stabilizes, leading to constant behavior across recurrences. Empirically, we discover that recurrent blocks learn stages of inference that closely mirror those of feedforward models, repeating these stages in depth with each iteration. We study how recurrent block size, input injection, and normalization influence the emergence and stability of these cyclic fixed points. We believe these findings help translate mechanistic insights into practical guidance for architectural design. --- ### ClawGuard: A Runtime Security Framework for Tool-Augmented LLM Agents Against Indirect Prompt Injection Authors: Wei Zhao, Zhe Li, Peixin Zhang, Jun Sun Published: 2026-04-13 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2604.11790v2 Tool-augmented Large Language Model (LLM) agents have demonstrated impressive capabilities in automating complex, multi-step real-world tasks, yet remain vulnerable to indirect prompt injection. Adversaries exploit this weakness by embedding malicious instructions within tool-returned content, which agents directly incorporate into their conversation history as trusted observations. To address these vulnerabilities, we introduce \textsc{ClawGuard}, a novel runtime security framework that enforces a user-confirmed rule set at every tool-call boundary, transforming unreliable alignment-dependent defense into a deterministic, auditable mechanism that intercepts adversarial tool calls before any real-world effect is produced. By automatically deriving task-specific access constraints from the user's stated objective prior to any external tool invocation, \textsc{ClawGuard} blocks all three injection pathways without model modification or infrastructure change. Experiments across five state-of-the-art language models on six injection benchmarks covering web, local, MCP, and skill channels, as well as three utility benchmarks covering OS, web, and code tasks, demonstrate that \textsc{ClawGuard} achieves robust protection against indirect prompt injection without compromising agent utility or introducing significant token overhead. This work establishes deterministic tool-call boundary enforcement as an effective defense mechanism for secure agentic AI systems. Code is publicly available at github.com/Claw-Guard/ClawGuard/. --- ### LMMs Meet Object-Centric Vision: Understanding, Segmentation, Editing and Generation Authors: Yuqian Yuan, Wenqiao Zhang, Juekai Lin, Yu Zhong, Mingjian Gao, Binhe Yu, Yunqi Cao, Wentong Li, Yueting Zhuang, Beng Chin Ooi Published: 2026-04-13 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.11789v2 Large Multimodal Models (LMMs) have achieved remarkable progress in general-purpose vision--language understanding, yet they remain limited in tasks requiring precise object-level grounding, fine-grained spatial reasoning, and controllable visual manipulation. In particular, existing systems often struggle to identify the correct instance, preserve object identity across interactions, and localize or modify designated regions with high precision. Object-centric vision provides a principled framework for addressing these challenges by promoting explicit representations and operations over visual entities, thereby extending multimodal systems from global scene understanding to object-level understanding, segmentation, editing, and generation. This paper presents a comprehensive review of recent advances at the convergence of LMMs and object-centric vision. We organize the literature into four major themes: object-centric visual understanding, object-centric referring segmentation, object-centric visual editing, and object-centric visual generation. We further summarize the key modeling paradigms, learning strategies, and evaluation protocols that support these capabilities. Finally, we discuss open challenges and future directions, including robust instance permanence, fine-grained spatial control, consistent multi-step interaction, unified cross-task modeling, and reliable benchmarking under distribution shift. We hope this paper provides a structured perspective on the development of scalable, precise, and trustworthy object-centric multimodal systems. --- ### HDR Video Generation via Latent Alignment with Logarithmic Encoding Authors: Naomi Ken Korem, Mohamed Oumoumad, Harel Cain, Matan Ben Yosef, Urska Jelercic, Ofir Bibi, Yaron Inger, Or Patashnik, Daniel Cohen-Or Published: 2026-04-13 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.11788v1 High dynamic range (HDR) imagery offers a rich and faithful representation of scene radiance, but remains challenging for generative models due to its mismatch with the bounded, perceptually compressed data on which these models are trained. A natural solution is to learn new representations for HDR, which introduces additional complexity and data requirements. In this work, we show that HDR generation can be achieved in a much simpler way by leveraging the strong visual priors already captured by pretrained generative models. We observe that a logarithmic encoding widely used in cinematic pipelines maps HDR imagery into a distribution that is naturally aligned with the latent space of these models, enabling direct adaptation via lightweight fine-tuning without retraining an encoder. To recover details that are not directly observable in the input, we further introduce a training strategy based on camera-mimicking degradations that encourages the model to infer missing high dynamic range content from its learned priors. Combining these insights, we demonstrate high-quality HDR video generation using a pretrained video model with minimal adaptation, achieving strong results across diverse scenes and challenging lighting conditions. Our results indicate that HDR, despite representing a fundamentally different image formation regime, can be handled effectively without redesigning generative models, provided that the representation is chosen to align with their learned priors. --- ### GenTac: Generative Modeling and Forecasting of Soccer Tactics Authors: Jiayuan Rao, Tianlin Gui, Haoning Wu, Yanfeng Wang, Weidi Xie Published: 2026-04-13 Categories: cs.AI, cs.MA arXiv: https://arxiv.org/abs/2604.11786v1 Modeling open-play soccer tactics is a formidable challenge due to the stochastic, multi-agent nature of the game. Existing computational approaches typically produce single, deterministic trajectory forecasts or focus on highly structured set-pieces, fundamentally failing to capture the inherent variance and branching possibilities of real-world match evolution. Here, we introduce GenTac, a diffusion-based generative framework that conceptualizes soccer tactics as a stochastic process over continuous multi-player trajectories and discrete semantic events. By learning the underlying distribution of player movements from historical tracking data, GenTac samples diverse, plausible, long-horizon future trajectories. The framework supports rich contextual conditioning, including opponent behavior, specific team or league playing styles, and strategic objectives, while grounding continuous spatial dynamics into a 15-class tactical event space. Extensive evaluations on our proposed benchmark, TacBench, demonstrate four key capabilities: (1) GenTac achieves high geometric accuracy while strictly preserving the collective structural consistency of the team; (2) it accurately simulates stylistic nuances, distinguishing between specific teams (e.g., Auckland FC) and leagues (e.g., A-League versus German leagues); (3) it enables controllable counterfactual simulations, demonstrably altering spatial control and expected threat metrics based on offensive or defensive guidance; and (4) it reliably anticipates future tactical outcomes directly from generated rollouts. Finally, we demonstrate that GenTac can be successfully trained to generalize to other dynamic team sports, including basketball, American football, and ice hockey. --- ### ClawGUI: A Unified Framework for Training, Evaluating, and Deploying GUI Agents Authors: Fei Tang, Zhiqiong Lu, Boxuan Zhang, Weiming Lu, Jun Xiao, Yueting Zhuang, Yongliang Shen Published: 2026-04-13 Categories: cs.LG, cs.AI, cs.CL, cs.CV arXiv: https://arxiv.org/abs/2604.11784v1 GUI agents drive applications through their visual interfaces instead of programmatic APIs, interacting with arbitrary software via taps, swipes, and keystrokes, reaching a long tail of applications that CLI-based agents cannot. Yet progress in this area is bottlenecked less by modeling capacity than by the absence of a coherent full-stack infrastructure: online RL training suffers from environment instability and closed pipelines, evaluation protocols drift silently across works, and trained agents rarely reach real users on real devices. We present \textbf{ClawGUI}, an open-source framework addressing these three gaps within a single harness. \textbf{ClawGUI-RL} provides the first open-source GUI agent RL infrastructure with validated support for both parallel virtual environments and real physical devices, integrating GiGPO with a Process Reward Model for dense step-level supervision. \textbf{ClawGUI-Eval} enforces a fully standardized evaluation pipeline across 6 benchmarks and 11+ models, achieving 95.8\% reproduction against official baselines. \textbf{ClawGUI-Agent} brings trained agents to Android, HarmonyOS, and iOS through 12+ chat platforms with hybrid CLI-GUI control and persistent personalized memory. Trained end to end within this pipeline, \textbf{ClawGUI-2B} achieves 17.1\% Success Rate on MobileWorld GUI-Only, outperforming the same-scale MAI-UI-2B baseline by 6.0\%. --- ### General365: Benchmarking General Reasoning in Large Language Models Across Diverse and Challenging Tasks Authors: Junlin Liu, Shengnan An, Shuang Zhou, Dan Ma, Shixiong Luo, Ying Xie, Yuan Zhang, Wenling Yuan, Yifan Zhou, Xiaoyu Li, Ziwen Wang, Xuezhi Cao et al. Published: 2026-04-13 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.11778v1 Contemporary large language models (LLMs) have demonstrated remarkable reasoning capabilities, particularly in specialized domains like mathematics and physics. However, their ability to generalize these reasoning skills to more general and broader contexts--often termed general reasoning--remains under-explored. Unlike domain-specific reasoning, general reasoning relies less on expert knowledge but still presents formidable reasoning challenges, such as complex constraints, nested logical branches, and semantic interference. To address this gap, we introduce General365, a benchmark specifically designed to assess general reasoning in LLMs. By restricting background knowledge to a K-12 level, General365 explicitly decouples reasoning from specialized expertise. The benchmark comprises 365 seed problems and 1,095 variant problems across eight categories, ensuring both high difficulty and diversity. Evaluations across 26 leading LLMs reveal that even the top-performing model achieves only 62.8% accuracy, in stark contrast to the near-perfect performances of LLMs in math and physics benchmarks. These results suggest that the reasoning abilities of current LLMs are heavily domain-dependent, leaving significant room for improvement in broader applications. We envision General365 as a catalyst for advancing LLM reasoning beyond domain-specific tasks toward robust, general-purpose real-world scenarios. Code, Dataset, and Leaderboard: https://general365.github.io --- ### Efficient KernelSHAP Explanations for Patch-based 3D Medical Image Segmentation Authors: Ricardo Coimbra Brioso, Giulio Sichili, Damiano Dei, Nicola Lambri, Pietro Mancosu, Marta Scorsetti, Daniele Loiacono Published: 2026-04-13 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2604.11775v1 Perturbation-based explainability methods such as KernelSHAP provide model-agnostic attributions but are typically impractical for patch-based 3D medical image segmentation due to the large number of coalition evaluations and the high cost of sliding-window inference. We present an efficient KernelSHAP framework for volumetric CT segmentation that restricts computation to a user-defined region of interest and its receptive-field support, and accelerates inference via patch logit caching, reusing baseline predictions for unaffected patches while preserving nnU-Net's fusion scheme. To enable clinically meaningful attributions, we compare three automatically generated feature abstractions within the receptive-field crop: whole-organ units, regular FCC supervoxels, and hybrid organ-aware supervoxels, and we study multiple aggregation/value functions targeting stabilizing evidence (TP/Dice/Soft Dice) or false-positive behavior. Experiments on whole-body CT segmentations show that caching substantially reduces redundant computation (with computational savings ranging from 15% to 30%) and that faithfulness and interpretability exhibit clear trade-offs: regular supervoxels often maximize perturbation-based metrics but lack anatomical alignment, whereas organ-aware units yield more clinically interpretable explanations and are particularly effective for highlighting false-positive drivers under normalized metrics. --- ### Autonomous Diffractometry Enabled by Visual Reinforcement Learning Authors: J. Oppliger, M. Stifter, A. Rüegg, I. Biało, L. Martinelli, P. G. Freeman, D. Prabhakaran, J. Zhao, Q. Wang, J. Chang Published: 2026-04-13 Categories: cs.LG, cond-mat.mtrl-sci, cs.CV arXiv: https://arxiv.org/abs/2604.11773v1 Automation underpins progress across scientific and industrial disciplines. Yet, automating tasks requiring interpretation of abstract visual information remain challenging. For example, crystal alignment strongly relies on humans with the ability to comprehend diffraction patterns. Here we introduce an autonomous system that aligns single crystals without access to crystallography and diffraction theory. Using a model-free reinforcement learning framework, an agent learns to identify and navigate towards high-symmetry orientations directly from Laue diffraction patterns. Despite the absence of human supervision, the agent develops human-like strategies to achieve time-efficient alignment across different crystal symmetry classes. With this, we provide a computational framework for intelligent diffractometers. As such, our approach advances the development of automated experimental workflows in materials science. --- ### MosaicMRI: A Diverse Dataset and Benchmark for Raw Musculoskeletal MRI Authors: Paula Arguello, Berk Tinaz, Mohammad Shahab Sepehri, Maryam Soltanolkotabi, Mahdi Soltanolkotabi Published: 2026-04-13 Categories: cs.CV, cs.LG, eess.SP, physics.med-ph, stat.ML arXiv: https://arxiv.org/abs/2604.11762v1 Deep learning underpins a wide range of applications in MRI, including reconstruction, artifact removal, and segmentation. However, progress has been driven largely by public datasets focused on brain and knee imaging, shaping how models are trained and evaluated. As a result, careful studies of the reliability of these models across diverse anatomical settings remain limited. In this work, we introduce MosaicMRI, a large and diverse collection of fully sampled raw musculoskeletal (MSK) MR measurements designed for training and evaluating machine-learning-based methods. MosaicMRI is the largest open-source raw MSK MRI dataset to date, comprising 2,671 volumes and 80,156 slices. The dataset offers substantial diversity in volume orientation (e.g., axial, sagittal), imaging contrasts (e.g., PD, T1, T2), anatomies (e.g., spine, knee, hip, ankle, and others), and numbers of acquisition coils. Using VarNet as a baseline for accelerated reconstruction task, we perform a comprehensive set of experiments to study scaling behavior with respect to both model capacity and dataset size. Interestingly, models trained on the combined anatomies significantly outperform anatomy-specific models in low-sample regimes, highlighting the benefits of anatomical diversity and the presence of exploitable cross-anatomical correlations. We further evaluate robustness and cross-anatomy generalization by training models on one anatomy (e.g., spine) and testing them on another (e.g., knee). Notably, we identify groups of body parts (e.g., foot and elbow) that generalize well with each other, and highlight that performance under domain shifts depends on both training set size, anatomy, and protocol-specific factors. --- ### Retrieval Is Not Enough: Why Organizational AI Needs Epistemic Infrastructure Authors: Federico Bottino, Carlo Ferrero, Nicholas Dosio, Pierfrancesco Beneventano Published: 2026-04-13 Categories: cs.AI arXiv: https://arxiv.org/abs/2604.11759v2 Organizational knowledge used by AI agents typically lacks epistemic structure: retrieval systems surface semantically relevant content without distinguishing binding decisions from abandoned hypotheses, contested claims from settled ones, or known facts from unresolved questions. We argue that the ceiling on organizational AI is not retrieval fidelity but \emph{epistemic} fidelity--the system's ability to represent commitment strength, contradiction status, and organizational ignorance as computable properties. We present OIDA, a framework that structures organizational knowledge as typed Knowledge Objects carrying epistemic class, importance scores with class-specific decay, and signed contradiction edges. The Knowledge Gravity Engine maintains scores deterministically with proved convergence guarantees (sufficient condition: max degree $< 7$; empirically robust to degree 43). OIDA introduces QUESTION-as-modeled-ignorance: a primitive with inverse decay that surfaces what an organization does \emph{not} know with increasing urgency--a mechanism absent from all surveyed systems. We describe the Epistemic Quality Score (EQS), a five-component evaluation methodology with explicit circularity analysis. In a controlled comparison ($n{=}10$ response pairs), OIDA's RAG condition (3,868 tokens) achieves EQS 0.530 vs.\ 0.848 for a full-context baseline (108,687 tokens); the $28.1\times$ token budget difference is the primary confound. The QUESTION mechanism is statistically validated (Fisher $p{=}0.0325$, OR$=21.0$). The formal properties are established; the decisive ablation at equal token budget (E4) is pre-registered and not yet run. --- ### StarVLA-$α$: Reducing Complexity in Vision-Language-Action Systems Authors: Jinhui Ye, Ning Gao, Senqiao Yang, Jinliang Zheng, Zixuan Wang, Yuxin Chen, Pengguang Chen, Yilun Chen, Shu Liu, Jiaya Jia Published: 2026-04-13 Categories: cs.RO, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2604.11757v2 Vision-Language-Action (VLA) models have recently emerged as a promising paradigm for building general-purpose robotic agents. However, the VLA landscape remains highly fragmented and complex: as existing approaches vary substantially in architectures, training data, embodiment configurations, and benchmark-specific engineering. In this work, we introduce StarVLA-$α$, a simple yet strong baseline designed to study VLA design choices under controlled conditions. StarVLA-$α$ deliberately minimizes architectural and pipeline complexity to reduce experimental confounders and enable systematic analysis. Specifically, we re-evaluate several key design axes, including action modeling strategies, robot-specific pretraining, and interface engineering. Across unified multi-benchmark training on LIBERO, SimplerEnv, RoboTwin, and RoboCasa, the same simple baseline remains highly competitive, indicating that a strong VLM backbone combined with minimal design is already sufficient to achieve strong performance without relying on additional architectural complexity or engineering tricks. Notably, our single generalist model outperforms $π_{0.5}$ by 20\% on the public real-world RoboChallenge benchmark. We expect StarVLA-$α$ to serve as a solid starting point for future research in the VLA regime. Code will be released at https://github.com/starVLA/starVLA. --- ### StarVLA-$α$: Reducing Complexity in Vision-Language-Action Systems Authors: Jinhui Ye, Ning Gao, Senqiao Yang, Jinliang Zheng, Zixuan Wang, Yuxin Chen, Pengguang Chen, Yilun Chen, Shu Liu, Jiaya Jia Published: 2026-04-13 Categories: cs.RO, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2604.11757v1 Vision-Language-Action (VLA) models have recently emerged as a promising paradigm for building general-purpose robotic agents. However, the VLA landscape remains highly fragmented and complex: as existing approaches vary substantially in architectures, training data, embodiment configurations, and benchmark-specific engineering. In this work, we introduce StarVLA-$α$, a simple yet strong baseline designed to study VLA design choices under controlled conditions. StarVLA-$α$ deliberately minimizes architectural and pipeline complexity to reduce experimental confounders and enable systematic analysis. Specifically, we re-evaluate several key design axes, including action modeling strategies, robot-specific pretraining, and interface engineering. Across unified multi-benchmark training on LIBERO, SimplerEnv, RoboTwin, and RoboCasa, the same simple baseline remains highly competitive, indicating that a strong VLM backbone combined with minimal design is already sufficient to achieve strong performance without relying on additional architectural complexity or engineering tricks. Notably, our single generalist model outperforms $π_{0.5}$ by 20\% on the public real-world RoboChallenge benchmark. We expect StarVLA-$α$ to serve as a solid starting point for future research in the VLA regime. Code will be released at https://github.com/starVLA/starVLA. --- ### Agentic Aggregation for Parallel Scaling of Long-Horizon Agentic Tasks Authors: Yoonsang Lee, Howard Yen, Xi Ye, Danqi Chen Published: 2026-04-13 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.11753v3 We study parallel test-time scaling for long-horizon agentic tasks such as agentic search and deep research, where multiple rollouts are generated in parallel and aggregated into a final response. While such scaling has proven effective for chain-of-thought reasoning, agentic tasks pose unique challenges: trajectories are long, multi-turn, and tool-augmented, and outputs are often open-ended. Aggregating only final answers discards rich information from trajectories, while concatenating all trajectories exceeds the model's context window. To address this, we propose AggAgent, an aggregation agent that treats parallel trajectories as an environment. We equip it with lightweight tools to inspect candidate solutions and search across trajectories, enabling it to navigate and synthesize information on demand. Across six benchmarks and three model families (GLM-4.7, Qwen3.5, MiniMax-M2.5), AggAgent outperforms all existing aggregation methods-by up to 5.3% absolute on average and 10.3% on two deep research tasks-while adding minimal overhead, as the aggregation cost remains bounded by a single agentic rollout. Our findings establish agentic aggregation as an effective and cost-efficient approach to parallel test-time scaling. --- ### Agentic Aggregation for Parallel Scaling of Long-Horizon Agentic Tasks Authors: Yoonsang Lee, Howard Yen, Xi Ye, Danqi Chen Published: 2026-04-13 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.11753v1 We study parallel test-time scaling for long-horizon agentic tasks such as agentic search and deep research, where multiple rollouts are generated in parallel and aggregated into a final response. While such scaling has proven effective for chain-of-thought reasoning, agentic tasks pose unique challenges: trajectories are long, multi-turn, and tool-augmented, and outputs are often open-ended. Aggregating only final answers discards rich information from trajectories, while concatenating all trajectories exceeds the model's context window. To address this, we propose AggAgent, an aggregation agent that treats parallel trajectories as an environment. We equip it with lightweight tools to inspect candidate solutions and search across trajectories, enabling it to navigate and synthesize information on demand. Across six benchmarks and three model families (GLM-4.7, Qwen3.5, MiniMax-M2.5), AggAgent outperforms all existing aggregation methods-by up to 5.3% absolute on average and 10.3% on two deep research tasks-while adding minimal overhead, as the aggregation cost remains bounded by a single agentic rollout. Our findings establish agentic aggregation as an effective and cost-efficient approach to parallel test-time scaling. --- ### Grounded World Model for Semantically Generalizable Planning Authors: Quanyi Li, Lan Feng, Haonan Zhang, Wuyang Li, Letian Wang, Alexandre Alahi, Harold Soh Published: 2026-04-13 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2604.11751v1 In Model Predictive Control (MPC), world models predict the future outcomes of various action proposals, which are then scored to guide the selection of the optimal action. For visuomotor MPC, the score function is a distance metric between a predicted image and a goal image, measured in the latent space of a pretrained vision encoder like DINO and JEPA. However, it is challenging to obtain the goal image in advance of the task execution, particularly in new environments. Additionally, conveying the goal through an image offers limited interactivity compared with natural language. In this work, we propose to learn a Grounded World Model (GWM) in a vision-language-aligned latent space. As a result, each proposed action is scored based on how close its future outcome is to the task instruction, reflected by the similarity of embeddings. This approach transforms the visuomotor MPC to a VLA that surpasses VLM-based VLAs in semantic generalization. On the proposed WISER benchmark, GWM-MPC achieves a 87% success rate on the test set comprising 288 tasks that feature unseen visual signals and referring expressions, yet remain solvable with motions demonstrated during training. In contrast, traditional VLAs achieve an average success rate of 22%, even though they overfit the training set with a 90% success rate. --- ### HistLens: Mapping Idea Change across Concepts and Corpora Authors: Yi Jing, Weiyun Qiu, Yihang Peng, Zhifang Sui Published: 2026-04-13 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.11749v1 Language change both reflects and shapes social processes, and the semantic evolution of foundational concepts provides a measurable trace of historical and social transformation. Despite recent advances in diachronic semantics and discourse analysis, existing computational approaches often (i) concentrate on a single concept or a single corpus, making findings difficult to compare across heterogeneous sources, and (ii) remain confined to surface lexical evidence, offering insufficient computational and interpretive granularity when concepts are expressed implicitly. We propose HistLens, a unified, SAE-based framework for multi-concept, multi-corpus conceptual-history analysis. The framework decomposes concept representations into interpretable features and tracks their activation dynamics over time and across sources, yielding comparable conceptual trajectories within a shared coordinate system. Experiments on long-span press corpora show that HistLens supports cross-concept, cross-corpus computation of patterns of idea evolution and enables implicit concept computation. By bridging conceptual modeling with interpretive needs, HistLens broadens the analytical perspectives and methodological repertoire available to social science and the humanities for diachronic text analysis. --- ### LangFlow: Continuous Diffusion Rivals Discrete in Language Modeling Authors: Yuxin Chen, Chumeng Liang, Hangke Sui, Ruihan Guo, Chaoran Cheng, Jiaxuan You, Ge Liu Published: 2026-04-13 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2604.11748v3 Continuous diffusion has been the foundation of high-fidelity, controllable, and few-step generation of many data modalities such as images. However, in language modeling, prior continuous diffusion language models (DLMs) lag behind discrete counterparts due to the sparse data space and the underexplored design space. In this work, we close this gap with LangFlow, the first continuous DLM to rival discrete diffusion, by connecting embedding-space DLMs to Flow Matching via Bregman divergence, alongside three key innovations: (1) we derive a novel ODE-based NLL bound for principled evaluation of continuous flow-based language models; (2) we propose an information-uniform principle for setting the noise schedule, which motivates a learnable noise scheduler based on a Gumbel distribution; and (3) we revise prior training protocols by incorporating self-conditioning, as we find it improves both likelihood and sample quality of embedding-space DLMs with effects substantially different from discrete diffusion. Putting everything together, LangFlow rivals top discrete DLMs on both the perplexity (PPL) and the generative perplexity (Gen. PPL), reaching a PPL of 30.0 on LM1B and 24.6 on OpenWebText. It even exceeds autoregressive baselines in zero-shot transfer on 4 out of 7 benchmarks. LangFlow provides the first clear evidence that continuous diffusion is a promising paradigm for language modeling. Homepage: https://github.com/nealchen2003/LangFlow --- ### Inferring Change Points in Regression via Sample Weighting Authors: Gabriel Arpino, Ramji Venkataramanan Published: 2026-04-13 Categories: stat.ME, math.ST, stat.ML arXiv: https://arxiv.org/abs/2604.11746v1 We study the problem of identifying change points in high-dimensional generalized linear models, and propose an approach based on sample-weighted empirical risk minimization. Our method, Weighted ERM, encodes priors on the change points via weights assigned to each sample, to obtain weighted versions of standard estimators such as M-estimators and maximum-likelihood estimators. Under mild assumptions on the data, we obtain a precise asymptotic characterization of the performance of our method for general Gaussian designs, in the high-dimensional limit where the number of samples and covariate dimension grow proportionally. We show how this characterization can be used to efficiently construct a posterior distribution over change points. Numerical experiments on both simulated and real data illustrate the efficacy of Weighted ERM compared to existing approaches, demonstrating that sample weights constructed with weakly informative priors can yield accurate change point estimators. Our method is implemented as an open-source package, weightederm, available in Python and R. --- ### KL Divergence Between Gaussians: A Step-by-Step Derivation for the Variational Autoencoder Objective Authors: Andrés Muñoz, Rodrigo Ramele Published: 2026-04-13 Categories: cs.LG arXiv: https://arxiv.org/abs/2604.11744v1 Kullback-Leibler (KL) divergence is a fundamental concept in information theory that quantifies the discrepancy between two probability distributions. In the context of Variational Autoencoders (VAEs), it serves as a central regularization term, imposing structure on the latent space and thereby enabling the model to exhibit generative capabilities. In this work, we present a detailed derivation of the closed-form expression for the KL divergence between Gaussian distributions, a case of particular importance in practical VAE implementations. Starting from the general definition for continuous random variables, we derive the expression for the univariate case and extend it to the multivariate setting under the assumption of diagonal covariance. Finally, we discuss the interpretation of each term in the resulting expression and its impact on the training dynamics of the model. --- ### Discourse Diversity in Multi-Turn Empathic Dialogue Authors: Hongli Zhan, Emma S. Gueorguieva, Javier Hernandez, Jina Suh, Desmond C. Ong, Junyi Jessy Li Published: 2026-04-13 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.11742v1 Large language models (LLMs) produce responses rated as highly empathic in single-turn settings (Ayers et al., 2023; Lee et al., 2024), yet they are also known to be formulaic generators that reuse the same lexical patterns, syntactic templates, and discourse structures across tasks (Jiang et al., 2025; Shaib et al., 2024; Namuduri et al., 2025). Less attention has been paid to whether this formulaicity extends to the level of discourse moves, i.e., what a response does for the person it is addressing. This question is especially consequential for empathic dialogue, where effective support demands not just a kind response at one moment but varied strategies as a conversation unfolds (Stiles et al., 1998). Indeed, prior work shows that LLMs reuse the same tactic sequences more than human supporters in single-turn settings (Gueorguieva et al., 2026). We extend this analysis to multi-turn conversations and find that the rigidity compounds: once a tactic appears in a supporter turn, LLMs reuse it in the next at nearly double the rate of humans (0.50-0.56 vs. 0.27). This pattern holds across LLMs serving as supporters in real emotional support conversations, and is invisible to standard similarity metrics. To address this gap, we introduce MINT (Multi-turn Inter-tactic Novelty Training), the first reinforcement learning framework to optimize discourse move diversity across multi-turn empathic dialogue. The best MINT variant combines an empathy quality reward with a cross-turn tactic novelty signal, improving aggregate empathy by 25.3% over vanilla across 1.7B and 4B models while reducing cross-turn discourse move repetition by 26.3% on the 4B model, surpassing all baselines including quality-only and token-level diversity methods on both measures. These results suggest that what current models lack is not empathy itself, but the ability to vary their discourse moves across a conversation. --- ### Collaborative Multi-Agent Scripts Generation for Enhancing Imperfect-Information Reasoning in Murder Mystery Games Authors: Keyang Zhong, Junlin Xie, Hefeng Wu, Haofeng Li, Guanbin Li Published: 2026-04-13 Categories: cs.AI arXiv: https://arxiv.org/abs/2604.11741v2 Vision-language models (VLMs) have shown impressive capabilities in perceptual tasks, yet they degrade in complex multi-hop reasoning under multiplayer game settings with imperfect and deceptive information. In this paper, we study a representative multiplayer task, Murder Mystery Games, which require inferring hidden truths based on partial clues provided by roles with different intentions. To address this challenge, we propose a collaborative multi-agent framework for evaluating and synthesizing high-quality, role-driven multiplayer game scripts, enabling fine-grained interaction patterns tailored to character identities (i.e., murderer vs. innocent). Our system generates rich multimodal contexts, including character backstories, visual and textual clues, and multi-hop reasoning chains, through coordinated agent interactions. We design a two-stage agent-monitored training strategy to enhance the reasoning ability of VLMs: (1) chain-of-thought based fine-tuning on curated and synthetic datasets that model uncertainty and deception; (2) GRPO-based reinforcement learning with agent-monitored reward shaping, encouraging the model to develop character-specific reasoning behaviors and effective multimodal multi-hop inference. Extensive experiments demonstrate that our method significantly boosts the performance of VLMs in narrative reasoning, hidden fact extraction, and deception-resilient understanding. Our contributions offer a scalable solution for training and evaluating VLMs under uncertain, adversarial, and socially complex conditions, laying the groundwork for future benchmarks in multimodal multi-hop reasoning under imperfect information. --- ### Collaborative Multi-Agent Scripts Generation for Enhancing Imperfect-Information Reasoning in Murder Mystery Games Authors: Keyang Zhong, Junlin Xie, Hefeng Wu, Haofeng Li, Guanbin Li Published: 2026-04-13 Categories: cs.AI arXiv: https://arxiv.org/abs/2604.11741v1 Vision-language models (VLMs) have shown impressive capabilities in perceptual tasks, yet they degrade in complex multi-hop reasoning under multiplayer game settings with imperfect and deceptive information. In this paper, we study a representative multiplayer task, Murder Mystery Games, which require inferring hidden truths based on partial clues provided by roles with different intentions. To address this challenge, we propose a collaborative multi-agent framework for evaluating and synthesizing high-quality, role-driven multiplayer game scripts, enabling fine-grained interaction patterns tailored to character identities (i.e., murderer vs. innocent). Our system generates rich multimodal contexts, including character backstories, visual and textual clues, and multi-hop reasoning chains, through coordinated agent interactions. We design a two-stage agent-monitored training strategy to enhance the reasoning ability of VLMs: (1) chain-of-thought based fine-tuning on curated and synthetic datasets that model uncertainty and deception; (2) GRPO-based reinforcement learning with agent-monitored reward shaping, encouraging the model to develop character-specific reasoning behaviors and effective multimodal multi-hop inference. Extensive experiments demonstrate that our method significantly boosts the performance of VLMs in narrative reasoning, hidden fact extraction, and deception-resilient understanding. Our contributions offer a scalable solution for training and evaluating VLMs under uncertain, adversarial, and socially complex conditions, laying the groundwork for future benchmarks in multimodal multi-hop reasoning under imperfect information. --- ### Learning Long-term Motion Embeddings for Efficient Kinematics Generation Authors: Nick Stracke, Kolja Bauer, Stefan Andreas Baumann, Miguel Angel Bautista, Josh Susskind, Björn Ommer Published: 2026-04-13 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.11737v1 Understanding and predicting motion is a fundamental component of visual intelligence. Although modern video models exhibit strong comprehension of scene dynamics, exploring multiple possible futures through full video synthesis remains prohibitively inefficient. We model scene dynamics orders of magnitude more efficiently by directly operating on a long-term motion embedding that is learned from large-scale trajectories obtained from tracker models. This enables efficient generation of long, realistic motions that fulfill goals specified via text prompts or spatial pokes. To achieve this, we first learn a highly compressed motion embedding with a temporal compression factor of 64x. In this space, we train a conditional flow-matching model to generate motion latents conditioned on task descriptions. The resulting motion distributions outperform those of both state-of-the-art video models and specialized task-specific approaches. --- ### SCORP: Scene-Consistent Multi-agent Diffusion Planning with Stable Online Reinforcement Post-Training for Cooperative Driving Authors: Haojie Bai, Aimin Li, Ruoyu Yao, Xiongwei Zhao, Tingting Zhang, Xing Zhang, Lin Gao, and Jun Ma Published: 2026-04-13 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2604.11734v3 Cooperative driving is a safety- and efficiency-critical task that requires the coordination of diverse, interaction-realistic multi-agent trajectories. Although existing diffusion-based methods can capture multimodal behaviors from demonstrations, they often exhibit weak scene consistency and poor alignment with closed-loop cooperative objectives. This makes post-training necessary for further improvement, yet achieving stable online post-training in reactive multi-agent environments remains challenging. In this paper, we propose SCORP, a scene-consistent multi-agent diffusion planner with stable online reinforcement learning (RL) post-training for cooperative driving. For pre-training, we develop a scene-conditioned multi-agent denoising architecture that couples inter-agent self-attention with a dual-path conditioning mechanism: cross-attention provides direct scene-information injection, while AdaLN-Zero enables additional flexible and stable conditional modulation, thereby improving the scene consistency and road adherence of joint trajectories. For post-training, we formulate a two-layer Markov decision process (MDP) that explicitly integrates the reverse denoising chain with policy-environment interaction. We further co-design dense, well-shaped planning rewards and variance-gated group-relative policy optimization (VG-GRPO) to mitigate advantage collapse and gradient instability during closed-loop training. Extensive experiments show that SCORP outperforms strong open-source baselines on WOMD, with 10.47%-28.26% and 1.70%-7.22% improvements in core safety and efficiency metrics, respectively. Moreover, compared with alternative post-training methods, SCORP delivers significant and consistent gains in both driving safety and traffic efficiency, highlighting stable and sustained advances in closed-loop cooperative driving. --- ### Endogenous Information in Routing Games: Memory-Constrained Equilibria, Recall Braess Paradoxes, and Memory Design Authors: Saad Alqithami Published: 2026-04-13 Categories: cs.GT, cs.AI, cs.IT arXiv: https://arxiv.org/abs/2604.11733v2 We study routing games in which travelers optimize over routes that are remembered or surfaced, rather than over a fixed exogenous action set. The paper develops a tractable design theory for endogenous recall and then connects it back to an explicit finite-memory micro model. At the micro level, each traveler carries a finite memory state, receives surfaced alternatives, chooses via a logit rule, and updates memory under a policy such as LRU. This yields a stationary Forgetful Wardrop Equilibrium (FWE); existence is proved under mild regularity, and uniqueness follows in a contraction regime for the reduced fixed-point map. The paper's main design layer is a stationary salience model that summarizes persistent memory and interface effects as route-specific weights. Salience-weighted stochastic user equilibrium is the unique minimizer of a strictly convex potential, which yields a clean optimization and implementability theory. In this layer we characterize governed implementability under ratio budgets and affine tying constraints, and derive constructive algorithms on parallel and series-parallel networks. The bridge between layers is exact for last-choice memory (B=1): the micro model is then equivalent to the salience model, so any interior salience vector can be realized by an appropriate surfacing policy. For larger memories, we develop an explicit LRU-to-TTL-to-salience approximation pipeline and add contraction-based bounds that translate surrogate-map error into fixed-point and welfare error. Finally, we define a Recall Braess Paradox, in which improving recall increases equilibrium delay without changing physical capacity, and show that it can arise on every two-terminal network with at least two distinct s-t paths. Targeted experiments support the approximation regime, governed-design predictions, and the computational advantages of the reduced layer. --- ### Endogenous Information in Routing Games: Memory-Constrained Equilibria, Recall Braess Paradoxes, and Memory Design Authors: Saad Alqithami Published: 2026-04-13 Categories: cs.GT, cs.AI, cs.IT arXiv: https://arxiv.org/abs/2604.11733v1 We study routing games in which travelers optimize over routes that are remembered or surfaced, rather than over a fixed exogenous action set. The paper develops a tractable design theory for endogenous recall and then connects it back to an explicit finite-memory micro model. At the micro level, each traveler carries a finite memory state, receives surfaced alternatives, chooses via a logit rule, and updates memory under a policy such as LRU. This yields a stationary Forgetful Wardrop Equilibrium (FWE); existence is proved under mild regularity, and uniqueness follows in a contraction regime for the reduced fixed-point map. The paper's main design layer is a stationary salience model that summarizes persistent memory and interface effects as route-specific weights. Salience-weighted stochastic user equilibrium is the unique minimizer of a strictly convex potential, which yields a clean optimization and implementability theory. In this layer we characterize governed implementability under ratio budgets and affine tying constraints, and derive constructive algorithms on parallel and series-parallel networks. The bridge between layers is exact for last-choice memory (B=1): the micro model is then equivalent to the salience model, so any interior salience vector can be realized by an appropriate surfacing policy. For larger memories, we develop an explicit LRU-to-TTL-to-salience approximation pipeline and add contraction-based bounds that translate surrogate-map error into fixed-point and welfare error. Finally, we define a Recall Braess Paradox, in which improving recall increases equilibrium delay without changing physical capacity, and show that it can arise on every two-terminal network with at least two distinct s-t paths. Targeted experiments support the approximation regime, governed-design predictions, and the computational advantages of the reduced layer. --- ### Nested Atoms Model with Application to Clustering Big Population-Scale Single-Cell Data Authors: Arhit Chakrabarti, Yang Ni, Yuchao Jiang, Bani K. Mallick Published: 2026-04-13 Categories: stat.ME, stat.AP, stat.ML arXiv: https://arxiv.org/abs/2604.11731v1 We consider the problem of clustering nested or hierarchical data, where observations are grouped and there are both group-level and observation-level variables. In our motivating OneK1K dataset, observations consist of single-cell RNA-sequencing (scRNA-seq) data from 982 individuals (groups), totaling 1.27 million cells (observations), along with individual-specific genotype data. This type of data would enable the identification of cell types and the investigation of how genetic variations among individuals influence differences in cell-type profiles. Our goal, therefore, is to jointly cluster cells and individuals to capture the heterogeneity across both levels using cell-specific gene expressions as well as individual-specific genotypes. However, existing grouped clustering methods do not incorporate group-level variables, thereby limiting their ability to capture the heterogeneity of genotypes in our motivating application. To address this, we propose the Nested Atoms Model (NAM), a new Bayesian nonparametric approach that enables the desired two-layered clustering, accounting for both group-level and observation-level variables. To scale NAM for high-dimensional data, we develop a fast variational Bayesian inference algorithm. Simulations show that NAM outperforms existing methods that ignore group-level variables. Applied to the OneK1K dataset, NAM identifies clusters of genetically similar individuals with homogeneous cell-type profiles. The resulting cell clusters align with known immune cell types based on differential gene expression, underscoring the ability of NAM to capture nested heterogeneity and provide biologically meaningful insights. --- ### Multimodal Ambivalence/Hesitancy Recognition in Videos for Personalized Digital Health Interventions Authors: Manuela González-González, Soufiane Belharbi, Muhammad Osama Zeeshan, Masoumeh Sharafi, Muhammad Haseeb Aslam, Lorenzo Sia, Nicolas Richet, Marco Pedersoli, Alessandro Lameiras Koerich, Simon L Bacon, Eric Granger Published: 2026-04-13 Categories: cs.CV, cs.HC, cs.LG arXiv: https://arxiv.org/abs/2604.11730v4 Using behavioural science, health interventions focus on behaviour change by providing a framework to help patients acquire and maintain healthy habits that improve medical outcomes. In-person interventions are costly and difficult to scale, especially in resource-limited regions. Digital health interventions offer a cost-effective approach, potentially supporting independent living and self-management. Automating such interventions, especially through machine learning, has recently gained considerable attention. Ambivalence and hesitancy (A/H) play a primary role for individuals to delay, avoid, or abandon health interventions. A/H are subtle and conflicting emotions that place a person in a state between positive and negative evaluations of a behaviour, or between acceptance and refusal to engage in it. They manifest as affective inconsistency across modalities or within a modality, such as language, facial, vocal expressions, and body language. While experts can be trained to recognize A/H, integrating them into digital health interventions is costly and less effective. Automatic A/H recognition is therefore critical for the personalization and cost-effectiveness of digital health interventions. Here, we explore the application of deep learning models for A/H recognition in videos, a multi-modal task by nature. In particular, this paper covers three learning setups: supervised learning, unsupervised domain adaptation for personalization, and zero-shot inference via large language models (LLMs). Our experiments are conducted on the unique and recently published BAH video dataset for A/H recognition. Our results show limited performance, suggesting that more adapted multi-modal models are required for accurate A/H recognition. Better methods for modeling spatio-temporal and multimodal fusion are necessary to leverage conflicts within/across modalities. ---