# Research Papers — 2026-04 · bundle page 50 of 233

> Full-text markdown of papers #2451–2500 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/50
[← Previous bundle](https://aipod.dev/papers/archive/2026-04/page/49.md)
[Next bundle →](https://aipod.dev/papers/archive/2026-04/page/51.md)

### Source-Modality Monitoring in Vision-Language Models

Authors: Etha Tianze Hua, Tian Yun, Ellie Pavlick
Published: 2026-04-23
Categories: cs.CL
arXiv: https://arxiv.org/abs/2604.22038v1

We define and investigate source-modality monitoring -- the ability of multimodal models to track and communicate the input source from which pieces of information originate. We consider source-modality monitoring as an instance of the more general binding problem, and evaluate the extent to which models exploit syntactic vs. semantic signals in order to bind words like image in a user-provided prompt to specific components of their input and context (i.e., actual images). Across experiments spanning 11 vision-language models (VLMs) performing target-modality information retrieval tasks, we find that both syntactic and semantic signals play an important role, but that the latter tend to outweigh the former in cases when modalities are highly distinct distributionally. We discuss the implications of these findings for model robustness, and in the context of increasingly multimodal agentic systems.

---

### EgoMAGIC- An Egocentric Video Field Medicine Dataset for Training Perception Algorithms

Authors: Brian VanVoorst, Nicholas Walczak, Christopher Gilleo, Charles Meissner, Fabio Felix, Iran Roman, Bea Steers, Claudio Silva, Yuhan Shen, Zijia Lu, Shih-Po Lee, Ehsan Elhamifar
Published: 2026-04-23
Categories: cs.CV, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2604.22036v1

This paper introduces EgoMAGIC (Medical Assistance, Guidance, Instruction, and Correction), an egocentric medical activity dataset collected as part of DARPA's Perceptually-enabled Task Guidance (PTG) program. This dataset comprises 3,355 videos of 50 medical tasks, with at least 50 labeled videos per task. The primary objective of the PTG program was to develop virtual assistants integrated into augmented reality headsets to assist users in performing complex tasks. To encourage exploration and research using this dataset, the medical training data has been released along with an action detection challenge focused on eight medical tasks. The majority of the videos were recorded using a head-mounted stereo camera with integrated audio. From this dataset, 40 YOLO models were trained using 1.95 million labels to detect 124 medical objects, providing a robust starting point for developers working on medical AI applications. In addition to introducing the dataset, this paper presents baseline results on action detection for the eight selected medical tasks across three models, with the best-performing method achieving average mAP 0.526. Although this paper primarily addresses action detection as the benchmark, the EgoMAGIC dataset is equally suitable for action recognition, object identification and detection, error detection, and other challenging computer vision tasks. The dataset is accessible via zenodo.org (DOI: 10.5281/zenodo.19239154).

---

### LTBs-KAN: Linear-Time B-splines Kolmogorov-Arnold Networks

Authors: Eduardo Said Merin-Martinez, Andres Mendez-Vazquez, Eduardo Rodriguez-Tello
Published: 2026-04-23
Categories: cs.LG, cs.CV, cs.NE
arXiv: https://arxiv.org/abs/2604.22034v1

Kolmogorov-Arnold Networks (KANs) are a recent neural network architecture offering an alternative to Multilayer Perceptrons (MLPs) with improved explainability and expressibility. However, KANs are significantly slower than MLPs due to the recursive nature of B-spline function computations, limiting their application. This work addresses these issues by proposing a novel base-spline Linear-Time B-splines Kolmogorov-Arnold Network (LTBs-KAN) with linear complexity. Unlike previous methods that rely on the Boor-Mansfield-Cox spline algorithm or other computationally intensive mathematical functions, our approach significantly reduces the computational burden. Additionally, we further reduce model's parameter through product-of-sums matrix factorization in the forward pass without sacrificing performance. Experiments on MNIST, Fashion-MNIST and CIFAR-10 demonstrate that LTBs-KAN achieves good time complexity and parameter reduction, when used as building architectural blocks, compared to other KAN implementations.

---

### Kernel Contracts: A Specification Language for ML Kernel Correctness Across Heterogeneous Silicon

Authors: Cooper Veit
Published: 2026-04-23
Categories: cs.LG, cs.PL
arXiv: https://arxiv.org/abs/2604.22032v1

Every ML kernel ships with an implicit contract about what it computes. People rarely write the contract down. When two kernels disagree -- when a matmul on AMD produces a different gradient than the same matmul on NVIDIA, when a fused attention kernel silently downcasts an accumulator, when an out-of-bounds access returns zero on one stack and garbage on another -- there is no formal artifact to arbitrate the dispute. Recent empirical work has measured the gap across silicon platforms, but none of it specifies the contract being violated. We present a specification language for kernel contracts. A contract has eight parts: identifier, scope, precondition, postcondition, tolerance, reference oracle, measurement protocol, and violation signature. We use it to state twelve contract classes covering precision, ordering, compiler-induced, and exceptional-value failure modes, each grounded in published empirical evidence. We require a three-state calibration: every contract must admit at least one reference-conforming implementation and at least one contract-violating implementation that passes basic functional tests. We apply the framework to three documented incidents -- Huawei Ascend silent precision coercion, Sakana AI CUDA Engineer reward hacking, AMD out-of-bounds silent acceptance -- and show that each informal diagnosis maps to a specific contract violation with a measurable signature. A kernel contract suite is a normative reference against which conformance can be graded, in the way that ISASecure grades industrial control systems against IEC 62443.

---

### Mochi: Aligning Pre-training and Inference for Efficient Graph Foundation Models via Meta-Learning

Authors: João Mattos, Arlei Silva
Published: 2026-04-23
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2604.22031v2

We propose Mochi, a Graph Foundation Model that addresses task unification and training efficiency by adopting a meta-learning based training framework. Prior models pre-train with reconstruction-based objectives such as link prediction, and assume that the resulting representations can be aligned with downstream tasks through a separate unification step such as class prototypes. We demonstrate through synthetic and real-world experiments that this procedure, while simple and intuitive, has limitations that directly affect downstream task performance. To address these limitations, Mochi pre-trains on few-shot episodes that mirror the downstream evaluation protocol, aligning the training objective with inference rather than relying on a post-hoc unification step. We show that Mochi, along with its more powerful variant Mochi++, achieves competitive or superior performance compared to existing Graph Foundation Models across 25 real-world graph datasets spanning node classification, link prediction, and graph classification, while requiring 8$\sim$27 times less training time than the strongest baseline.

---

### Shared Lexical Task Representations Explain Behavioral Variability In LLMs

Authors: Zhuonan Yang, Jacob Xiaochen Li, Francisco Piedrahita Velez, Eric Todd, David Bau, Michael L. Littman, Stephen H. Bach, Ellie Pavlick
Published: 2026-04-23
Categories: cs.CL, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2604.22027v2

One of the most common complaints about large language models (LLMs) is their prompt sensitivity -- that is, the fact that their ability to perform a task or provide a correct answer to a question can depend unpredictably on the way the question is posed. We investigate this variation by comparing two very different but commonly-used styles of prompting: instruction-based prompts, which describe the task in natural language, and example-based prompts, which provide in-context few-shot demonstration pairs to illustrate the task. We find that, despite large variation in performance as a function of the prompt, the model engages some common underlying mechanisms across different prompts of a task. Specifically, we identify task-specific attention heads whose outputs literally describe the task -- which we dub lexical task heads -- and show that these heads are shared across prompting styles and trigger subsequent answer production. We further find that behavioral variation between prompts can be explained by the degree to which these heads are activated, and that failures are at least sometimes due to competing task representations that dilute the signal of the target task. Our results together present an increasingly clear picture of how LLMs' internal representations can explain behavior that otherwise seems idiosyncratic to users and developers.

---

### Rethinking Publication: A Certification Framework for AI-Enabled Research

Authors: Yang Lu, Rabimba Karanjai, Lei Xu, Weidong Shi
Published: 2026-04-23
Categories: cs.AI, cs.CY, cs.DL
arXiv: https://arxiv.org/abs/2604.22026v2

AI research pipelines can now generate academic work that may satisfy existing peer review standards for quality, novelty, and methodological rigor. However, the publication system was built around the assumption that research is produced by human authors. It therefore lacks a clear way to evaluate work when the knowledge claim may be valid but the producer is partly or fully automated. This paper proposes a two-layer certification framework for AI-generated research. The first layer evaluates whether the knowledge claim is sound. The second layer evaluates the level of human contribution. This separation allows journals and conferences to assess pipeline-generated work more consistently without creating new institutions. The framework uses normative analysis, conceptual design, and dry-run validation against representative submission cases. It classifies human contribution into three categories: Category A, where the work is reachable by an automated pipeline; Category B, where human direction is required at identifiable stages; and Category C, where the work goes beyond current pipeline capability, especially at the problem-formulation stage. The paper also proposes dedicated benchmark slots for fully disclosed automated research. These slots would provide a transparent publication path and help reviewers calibrate judgments over time. The key argument is that publication has historically certified two things at once: that the knowledge is valid and that a human produced it. AI research pipelines separate these two claims. By decoupling knowledge certification from authorship attribution, the proposed framework responds to a structural change already underway. It can be implemented within existing editorial systems, works even when attribution is uncertain, and recognizes human frontier contribution based on epistemic value rather than human origin alone.

---

### Foundation models for discovering robust biomarkers of neurological disorders from dynamic functional connectivity

Authors: Deepank Girish, Yi Hao Chan, Sukrit Gupta, Jing Xia, Jagath C. Rajapakse
Published: 2026-04-23
Categories: q-bio.NC, cs.AI, cs.LG, eess.SP
arXiv: https://arxiv.org/abs/2604.22018v1

Several brain foundation models (FM) have recently been proposed to predict brain disorders by modelling dynamic functional connectivity (FC). While they demonstrate remarkable model performance and zero- or few-shot generalization, the salient features identified as potential biomarkers are yet to be thoroughly evaluated. We propose RE-CONFIRM, a framework for evaluating the robustness of potential biomarker candidates elucidated by deep learning (DL) models including FMs. From experiments on five large datasets of Autism Spectrum Disorder (ASD), Attention-deficit Hyperactivity Disorder (ADHD), and Alzheimer's Disease (AD), we found that although commonly used performance metrics provide an intuitive assessment of model predictions, they are insufficient for evaluating the robustness of biomarkers identified by these models. RE-CONFIRM metrics revealed that simply finetuning FMs leads to models that fail to capture regional hubs effectively, even in disorders where hubs are known to be implicated, such as ASD and ADHD. In view of this, we propose Hub-LoRA (Low-Rank Adaptation) as a fine-tuning technique that enables FMs to not only outperform customised DL models but also produce neurobiologically faithful biomarkers supported by meta-analyses. RE-CONFIRM is generalizable and can be easily applied to ascertain the robustness of DL models trained on functional MRI datasets. Code is available at: https://github.com/SCSE-Biomedical-Computing-Group/RE-CONFIRM.

---

### Hierarchical Probabilistic Principal Component Analysis of Longitudinal Data

Authors: Xinyu Zhang, Ameer Qaqish, D. Y. Lin, Didong Li
Published: 2026-04-23
Categories: stat.ME, stat.AP, stat.ML
arXiv: https://arxiv.org/abs/2604.22015v1

In many longitudinal studies, a large number of variables are measured repeatedly over time, with substantial missing data. Existing methods, such as probabilistic principal component analysis (PPCA), are ill-equipped to handle such incomplete, high-dimensional longitudinal data, as they fail to account for the nested sources of variation and temporal dependency inherent in repeated measures. We introduce hierarchical probabilistic principal component analysis (HPPCA), a two-level probabilistic factor model that explicitly separates between-subject variance from time-varying within-subject dynamics. The within-subject latent factors are modeled by a Gaussian process. We develop an EM algorithm to handle missing data and flexible covariance kernels, accelerated by computationally efficient initializers. Simulation studies demonstrated that HPPCA robustly recovers model parameters subspaces and substantially outperforms both standard PPCA and multivariate functional PCA in imputation accuracy, even under heavy missingness and model misspecification. An application to the long COVID symptoms in the Researching COVID to Enhance Recovery adult cohort revealed that HPPCA effectively captured the data's hierarchical structure and its learned features significantly improved the prediction of clinical outcomes and the recovery of masked clinical records compared to exisiting methods.

---

### Null-Space Flow Matching for MIMO Channel Estimation in Latency-Constrained Systems

Authors: Junjie Zhao, Guangming Liang, Xiaonan Liu, Dongzhu Liu
Published: 2026-04-23
Categories: cs.IT, cs.LG, eess.SP
arXiv: https://arxiv.org/abs/2604.22005v2

Accurate yet low-latency channel state information (CSI) acquisition is essential for multiple-input multiple-output (MIMO) communication systems. While advanced deep generative models, such as score-based and diffusion models, enable high-fidelity CSI reconstruction from limited pilot observations, they often suffer from high inference latency. To achieve accurate CSI estimation under stringent latency constraints, this paper proposes a null-space flow matching (FM) framework that leverages a range-null space decomposition to separate observation-informed and underdetermined channel components. Specifically, the pilot observations are used to regulate the observable range-space channel component, while an FM-based generative prior primarily resolves the ambiguous null-space degrees of freedom through iterative refinement. To further improve the robustness and efficiency of the proposed framework, we introduce a noise-aware adaptive correction strategy to suppress channel noise on the refinement trajectory, along with a power-law time schedule to better allocate the limited number of refinement steps. Experimental results demonstrate that our method achieves competitive normalized mean square error (NMSE) performance even under a strict latency budget of around 3 ms, while delivering a superior accuracy-latency tradeoff compared with both model-based and generative baselines.

---

### Null-Space Flow Matching for MIMO Channel Estimation in Latency-Constrained Systems

Authors: Junjie Zhao, Guangming Liang, Dongzhu Liu, Xiaonan Liu
Published: 2026-04-23
Categories: cs.IT, cs.LG, eess.SP
arXiv: https://arxiv.org/abs/2604.22005v1

Accurate yet low-latency channel state information (CSI) acquisition is essential for multiple-input multiple-output (MIMO) communication systems. While advanced deep generative models, such as score-based and diffusion models, enable high-fidelity CSI reconstruction from limited pilot observations, they often suffer from high inference latency. To achieve accurate CSI estimation under stringent latency constraints, this paper proposes a null-space flow matching (FM) framework that decomposes pilot-limited MIMO channel estimation into a range-space reconstruction problem and a null-space generation problem. Specifically, the range-space component of the channel is directly recovered from noisy pilot observations, while only the ambiguous null-space component is iteratively refined using an FM-based generative prior. To further improve the robustness of the proposed framework, we introduce a power-law time schedule to better allocate the limited number of refinement steps, along with a noise-aware adaptive correction strategy to suppress channel noise on the refinement trajectory. Experimental results demonstrate that our method achieves a competitive normalized mean square error (NMSE) even under a strict latency budget of around 3 ms, while delivering superior estimation accuracy and faster inference than both model-based and generative baselines.

---

### When Cow Urine Cures Constipation on YouTube: Limits of LLMs in Detecting Culture-specific Health Misinformation

Authors: Anamta Khan, Ratna Kandala, Deepti, Sheza Munir, Joyojeet Pal
Published: 2026-04-23
Categories: cs.CL
arXiv: https://arxiv.org/abs/2604.22002v1

Social media platforms have become primary channels for health information in the Global South. Using gomutra (cow urine) discourse on YouTube in India as a case study, we present a post-facto Large Language Model (LLM)-assisted discourse analysis of 30 multilingual transcripts showing that promotional content blends sacred traditional language with pseudo-scientific claims in ways that sophisticated debunking content itself mirrors, creating a rhetorical register that LLMs, trained predominantly on Western corpora, are systematically ill-equipped to analyse. Varying prompt tone across three LLMs (GPT-4o, Gemini 2.5 Pro, DeepSeek-V3.1), we find that culturally embedded health misinformation does not look like ordinary misinformation, and this cultural obfuscation extends to gendered rhetoric and prompt design, compounding analytical unreliability. Our findings argue that cultural competency in LLM-assisted discourse analysis cannot be retrofitted through prompt engineering alone.

---

### L-System Genetic Encoding for Scalable Neural Network Evolution: A Comparison with Direct Matrix Encoding

Authors: Alexander Stuy, Nodin Weddington
Published: 2026-04-23
Categories: cs.NE
arXiv: https://arxiv.org/abs/2604.22000v2

An artificial world of barriers and plains scattered with food is used to test the feasibility of using genetic algorithms to optimize hebbian neural networks to perform on problems without apriori knowledge of the problem domain. A formal L-System based genetic alphabet for neural networks, titled Lsys, and a neural network genetic modeling tool titled Wp1hgn are introduced. Lsys and Matrix neural network topology genetic encoding methods are compared across 24 experimental runs. Lsys encoding achieved a mean maximum food count of 3802 +- 197 at generation 1000 across 8 runs with varied parameters, compared to 1388 +- 610 for Matrix encoding, a 2.74x performance advantage with an 8.5-fold improvement in consistency as measured by coefficient of variation (5.2% vs 44.0%). All 8 Lsys populations successfully learned to navigate the environment, while 4 of 8 Matrix populations failed to achieve competitive performance at any point during 1000 generations. When transferred to a novel maze environment, Lsys populations demonstrated immediate robust generalization, achieving a mean maximum food count of 2455 +- 176 compared to 422 +- 212 for Matrix populations, a 5.82x advantage that exceeded the training world performance gap. A MatrixLSG control condition, in which initial populations were generated using Lsys genotypes and then evolved using Matrix operators, demonstrated that the performance advantage of Lsys encoding derives primarily from the genetic algorithm operating on the compressed symbolic Lsys alphabet throughout evolution rather than from initial population structure. Lsys encoding is shown to provide faster convergence, higher peak performance, dramatically greater reliability, and superior generalization to novel environments compared to Matrix encoding across all experimental conditions tested.

---

### Universal Transformers Need Memory: Depth-State Trade-offs in Adaptive Recursive Reasoning

Authors: Grigory Sapunov
Published: 2026-04-23
Categories: cs.LG, cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2604.21999v3

We study learned memory tokens as a computational scratchpad for a single-block Universal Transformer with Adaptive Computation Time (ACT) on Sudoku-Extreme, a combinatorial reasoning benchmark. Memory tokens are empirically necessary: no configuration without them reaches non-trivial performance. The optimal count has a sharp lower threshold (T=0 always fails, T=8 reliably succeeds) followed by a stable plateau (T=8-32, 57.4% +/- 0.7% exact-match) and a dilution boundary at T=64. Under halt-side pressure (lambda warmup), mean halt drops monotonically with memory size across the plateau (from 11.6 at T=8 to 8.3 at T=64), showing that memory tokens and ponder depth substitute as resources at fixed accuracy. We also identify a router initialization trap that causes the majority of training runs to fail: both default zero-bias and Graves' recommended positive bias settle into a shallow halt equilibrium the model cannot escape. Inverting the bias to -3 ("deep start") eliminates the failure mode, and ablation shows the trap is inherent to ACT initialization rather than an artifact of our architecture. With reliable training, ACT yields an order of magnitude lower seed variance than fixed-depth processing (+/-0.7 vs +/-9.3 pp); lambda warmup recovers 34% of compute at matched accuracy; and attention heads specialize into memory readers, constraint propagators, and integrators across recursive depth. Code: https://github.com/che-shr-cat/utm-jax.

---

### Contrast-Space Projection for Network Meta-Analysis: An Exact and Invariant Study-Based Decomposition of Direct and Indirect Contributions

Authors: Chong Wang, Yanqi Zhang, Zhezhen Jin, Annette O'Connor
Published: 2026-04-23
Categories: stat.ME, stat.AP, stat.CO, stat.ML
arXiv: https://arxiv.org/abs/2604.21994v2

Network meta-analysis (NMA) combines direct and indirect comparisons across a treatment network, but exact contribution decompositions that reproduce NMA estimates are lacking, especially for multi-arm trials with correlated contrasts. We develop a contrast-space projection formulation of NMA that expresses the estimator as a linear mapping of observed pairwise contrasts onto the consistency-constrained contrast space. Building on this representation, we define direct and indirect evidence through a canonical within-study reduction that removes algebraic redundancy and yields a unique, invariant study-level decomposition. The resulting covariance-aware weights exactly reconstruct the NMA estimator and can be further resolved into indirect path-level components. Under fixed effects, the same projection also represents the generalized Cochran Q decomposition into within-design heterogeneity and between-design inconsistency. The framework yields diagnostic and graphical tools, including forest plots, tension plots, and path-based visualizations. Applications to empirical networks illustrate how the approach provides a reproducible and interpretable account of evidence contributions in NMA.

---

### Contrast-Space Projection for Network Meta-Analysis: An Exact and Invariant Study-Based Decomposition of Direct and Indirect Contributions

Authors: Chong Wang, Yanqi Zhang, Zhezhen Jin, Annette O'Connor
Published: 2026-04-23
Categories: stat.ME, stat.AP, stat.CO, stat.ML
arXiv: https://arxiv.org/abs/2604.21994v1

Network meta-analysis (NMA) combines direct and indirect comparisons across a connected treatment network to estimate relative treatment effects. However, there is a lack of exact contribution decompositions that reproduce NMA estimates, particularly in the presence of multi-arm trials that induce within-study correlations. We address this reproducibility gap by developing a contrast-space projection formulation of NMA. Working in the space of all estimable pairwise treatment contrasts, we express the NMA estimator as an explicit linear mapping of the observed contrasts onto the consistency-constrained contrast space induced by orthogonal projection. Building on this representation, we introduce a rigorous study-based definition of direct and indirect evidence through a canonical within-study reduction that removes algebraic redundancy and yields a unique, invariant decomposition. This leads to exact covariance-aware decompositions of the NMA estimator into study-level direct and indirect contributions, with indirect evidence further resolved into path-level components. The resulting weights are directly analogous to inverse-variance weights in pairwise meta-analysis and enable, to our knowledge, the first forest-plot representation that exactly reconstructs the NMA estimator. The framework also yields projection-based diagnostic and graphical tools, including forest plots, tension plots, and path-based visualizations. Applications to empirical datasets demonstrate how the proposed approach provides a reproducible and interpretable framework for understanding evidence contributions in network meta-analysis, supporting transparent interpretation and reporting.

---

### When Quotes Crumble: Detecting Transient Mechanical Liquidity Erosion in Limit Order Books

Authors: Haohan Xu, Jason Bohne, Pawel Polak, Yurij Baransky, Ajay Alva, Violetta Fedotova, Gary Kazantsev, David Rosenberg
Published: 2026-04-23
Categories: cs.LG
arXiv: https://arxiv.org/abs/2604.21993v1

We study the detection of transient liquidity erosion ("crumbling quotes") in electronic limit order books, where observable quote deterioration may reflect either mechanical liquidity withdrawal or informational repricing. Using the ABIDES agent-based simulator, we construct a multi-agent environment in which crumbling emerges from stochastic regime switches in a market maker, providing time-resolved ground truth unavailable in real market data. We develop a detection pipeline that identifies mechanically driven quote erosion using order book features, and train a neural model to produce calibrated crumbling probabilities. Experiments demonstrate that the proposed framework reliably identifies crumbling events against agent-level ground truth, with the neural model achieving +36% AUC improvement over rule-based baselines and robust performance across normal, high-volatility, bull, and bear market conditions. Ablation studies on temporal features and varying the dependence structure of the ground-truth mechanism confirm that the framework generalizes across both independent and autocorrelated liquidity withdrawal dynamics.

---

### Multi-Task Optimization over Networks of Tasks

Authors: Julian Hatzky, Thomas Bartz-Beielstein, A. E. Eiben, Anil Yaman
Published: 2026-04-23
Categories: cs.LG, cs.AI, cs.NE
arXiv: https://arxiv.org/abs/2604.21991v2

Multi-task optimization is a powerful approach for solving a large number of tasks in parallel. However, existing algorithms face distinct limitations: Population-based methods scale poorly and remain underexplored for large task sets. Approaches that do scale beyond a thousand tasks are mostly MAP-Elites variants and rely on a fixed, discretized archive that disregards the topology of the task space. We introduce MONET (Multi-Task Optimization over Networks of Tasks), a multi-task optimization algorithm that models the task space as a graph: tasks are nodes, and edges connect tasks in the task parameter space. This representation enables knowledge transfer between tasks and remains tractable for high-dimensional problems while exploiting the topology of the task space. MONET combines social learning, which generates candidates from neighboring nodes via crossover, with individual learning, which refines a node's own solution independently via mutation. We evaluate MONET on four domains (archery, arm, and cartpole with 5,000 tasks each; hexapod with 2,000 tasks) and show that it matches or exceeds the performance of existing MAP-Elites-based baselines across all four domains.

---

### Model Predictive Control of Hybrid Dynamical Systems

Authors: Ricardo G. Sanfelice, Berk Altin
Published: 2026-04-23
Categories: math.OC, cs.AI, cs.RO, eess.SY, math.DS
arXiv: https://arxiv.org/abs/2604.21989v2

The problem of controlling hybrid dynamical systems using model predictive control (MPC) is formulated and sufficient conditions for asymptotic stability of a set are provided. Hybrid dynamical systems are modeled in terms of hybrid equations, involving a differential equation and a difference equation with inputs and constraints. The proposed hybrid MPC algorithm uses a suitable prediction and control horizon construction inspired by hybrid time domains. Structural properties of the hybrid optimization problem, its feasible set, and its value function are provided. Checkable conditions to guarantee asymptotic stability of a set are provided. These conditions are given in terms of properties on the stage cost, terminal cost, and the existence of static state-feedback laws, related through a control Lyapunov function condition. Examples illustrate the results throughout the paper.

---

### Model Predictive Control of Hybrid Dynamical Systems

Authors: Ricardo G. Sanfelice, Berk Altin
Published: 2026-04-23
Categories: math.OC, cs.AI, cs.RO, eess.SY, math.DS
arXiv: https://arxiv.org/abs/2604.21989v1

The problem of controlling hybrid dynamical systems using model predictive control (MPC) is formulated and sufficient conditions for asymptotic stability of a set are provided. Hybrid dynamical systems are modeled in terms of hybrid equations, involving a differential equation and a difference equation with inputs and constraints. The proposed hybrid MPC algorithm uses a suitable prediction and control horizon construction inspired by hybrid time domains. Structural properties of the hybrid optimization problem, its feasible set, and its value function are provided. Checkable conditions to guarantee asymptotic stability of a set are provided. These conditions are given in terms of properties on the stage cost, terminal cost, and the existence of static state-feedback laws, related through a control Lyapunov function condition. Examples illustrate the results throughout the paper.

---

### Soft Anisotropic Diagrams for Differentiable Image Representation

Authors: Laki Iinbor, Zhiyang Dou, Wojciech Matusik
Published: 2026-04-23
Categories: cs.CV
arXiv: https://arxiv.org/abs/2604.21984v2

We introduce Soft Anisotropic Diagrams (SAD), an explicit and differentiable image representation parameterized by a set of adaptive sites in the image plane. In SAD, each site specifies an anisotropic metric and an additively weighted distance score, and we compute pixel colors as a softmax blend over a small per-pixel top-K subset of sites. We induce a soft anisotropic additively weighted Voronoi partition (i.e., an Apollonius diagram) with learnable per-site temperatures, preserving informative gradients while allowing clear, content-aligned boundaries and explicit ownership. Such a formulation enables efficient rendering by maintaining a per-query top-K map that approximates nearest neighbors under the same shading score, allowing GPU-friendly, fixed-size local computation. We update this list using our top-K propagation scheme inspired by jump flooding, augmented with stochastic injection to provide probabilistic global coverage. Training follows a GPU-first pipeline with gradient-weighted initialization, Adam optimization, and adaptive budget control through densification and pruning. Across standard benchmarks, SAD consistently outperforms Image-GS and Instant-NGP at matched bitrate. On Kodak, SAD reaches 46.0 dB PSNR with 2.2 s encoding time (vs. 28 s for Image-GS), and delivers 4-19 times end-to-end training speedups over state-of-the-art baselines. We demonstrate the effectiveness of SAD by showcasing the seamless integration with differentiable pipelines for forward and inverse problems, efficiency of fast random access, and compact storage.

---

### Forecasting Solar Energy Using a Single Image

Authors: Jeremy Klotz, Shree K. Nayar
Published: 2026-04-23
Categories: cs.CV
arXiv: https://arxiv.org/abs/2604.21982v1

Solar panels are increasingly deployed in cities on rooftops, walls, and urban infrastructure. Although the panel costs have fallen in recent years, the soft costs of installing them have not. These soft costs include assessing the illumination (irradiance) of a panel, which is typically performed using a 3D model that fails to capture small nearby structures that impact the irradiance. Our approach uses a single image taken at the panel's location to forecast its irradiance at any time in the future. We use visual cues in the image to find the camera's orientation and the portion of the sky visible to the panel in order to forecast the irradiance due to the sun and the sky. In addition, we show that the irradiance due to reflections from nearby buildings varies smoothly over time and can be forecasted from the image. This approach enables assessing the solar energy potential of any surface and forecasting the temporal variation of a panel's irradiance. We validate our approach using real irradiance measurements in urban canyons. We show that our approach often yields more accurate irradiance forecasts compared to conventional irradiance-based transposition methods and 3D model-based simulations. We also show that a single spherical image can be used to find the best fixed orientation of a panel. Finally, we present Solaris, a device to capture the image seen by a panel in a variety of urban settings.

---

### Read the Paper, Write the Code: Agentic Reproduction of Social-Science Results

Authors: Benjamin Kohler, David Zollikofer, Johanna Einsiedler, Alexander Hoyle, Elliott Ash
Published: 2026-04-23
Categories: cs.AI
arXiv: https://arxiv.org/abs/2604.21965v1

Recent work has used LLM agents to reproduce empirical social science results with access to both the data and code. We broaden this scope by asking: Can they reproduce results given only a paper's methods description and original data? We develop an agentic reproduction system that extracts structured methods descriptions from papers, runs reimplementations under strict information isolation -- agents never see the original code, results, or paper -- and enables deterministic, cell-level comparison of reproduced outputs to the original results. An error attribution step traces discrepancies through the system chain to identify root causes. Evaluating four agent scaffolds and four LLMs on 48 papers with human-verified reproducibility, we find that agents can largely recover published results, but performance varies substantially between models, scaffolds, and papers. Root cause analysis reveals that failures stem both from agent errors and from underspecification in the papers themselves.

---

### A general optimization solver based on OP-to-MaxSAT reduction

Authors: Yuxin Zhao, Han Huang, Zhifeng Hao
Published: 2026-04-23
Categories: cs.LO, cs.AI, cs.SC
arXiv: https://arxiv.org/abs/2604.21961v1

Optimization problems are fundamental in diverse fields, such as engineering, economics, and scientific computing. However, current algorithms are mostly designed for specific problem types and exhibit limited generality in solving multiple types of optimization problems. To enhance generality, we propose an automated reduction method named OP-to-MaxSAT reduction and a general optimization solver based on OP-to-MaxSAT reduction (GORED). GORED unifies the solving of multiple types of optimization problems by reducing the problems from optimization problems to MaxSAT instances in polynomial time and solving them using the state-of-the-art MaxSAT solver. The generality and solution quality of GORED are validated through experiments on 136 instances across 11 types of optimization problems. Experimental results demonstrate that GORED not only successfully solves a wide range of optimization problems but also yields solutions comparable in quality to those from existing methods, with no statistically significant differences observed. By introducing automated reduction, this work shifts the paradigm of optimization solvers from designing specialized algorithms for each problem type to employing a single algorithm for diverse problems. As a result, advances in this single algorithm can now drive progress in a wide range of optimization problems across various domains.

---

### Conditional Diffusion Posterior Alignment for Sparse-View CT Reconstruction

Authors: Luis Barba, Johannes Kirschner, Benjamin Bejar
Published: 2026-04-23
Categories: eess.IV, cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2604.21960v3

Computed Tomography (CT) is a widely used imaging modality in medical and industrial applications. To limit radiation exposure and measurement time, there is a growing interest in sparse-view CT, where the number of projection views is significantly reduced. Deep neural networks have shown great promise in improving reconstruction quality in sparse-view CT, especially generative diffusion models. However, these methods struggle to scale to large 3D volumes due to several reasons: (i) the high memory and computational requirements of 3D models, (ii) the lack of large 3D training datasets, and (iii) the inconsistencies across slices when using 2D models independently on each slice. We overcome these limitations and scale diffusion-based sparse-view CT reconstruction to large 3D volumes by combining conditional diffusion with explicit data consistency. We propose Conditional Diffusion Posterior Alignment (CDPA) to enable scalable 3D sparse-view CT reconstruction. A 2D U-Net diffusion model is conditioned on an initial 3D reconstruction to improve inter-slice consistency, combined with data-consistency alignment to match measured projections. Experiments on synthetic and real Cone Beam CT (CBCT) data show state-of-the-art performance, with ablations that confirm the synergistic effects of the proposed pipeline. Finally, we show that the same principles also strengthen fast denoising U-Nets, yielding near-diffusion quality at a fraction of the computational cost.

---

### Conditional Diffusion Posterior Alignment for Sparse-View CT Reconstruction

Authors: Luis Barba, Johannes Kirschner, Benjamin Bejar
Published: 2026-04-23
Categories: eess.IV, cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2604.21960v2

Computed Tomography (CT) is a widely used imaging modality in medical and industrial applications. To limit radiation exposure and measurement time, there is a growing interest in sparse-view CT, where the number of projection views is significantly reduced. Deep neural networks have shown great promise in improving reconstruction quality in sparse-view CT, especially generative diffusion models. However, these methods struggle to scale to large 3D volumes due to several reasons: (i) the high memory and computational requirements of 3D models, (ii) the lack of large 3D training datasets, and (iii) the inconsistencies across slices when using 2D models independently on each slice. We overcome these limitations and scale diffusion-based sparse-view CT reconstruction to large 3D volumes by combining conditional diffusion with explicit data consistency. We propose Conditional Diffusion Posterior Alignment (CDPA) to enable scalable 3D sparse-view CT reconstruction. A 2D U-Net diffusion model is conditioned on an initial 3D reconstruction to improve inter-slice consistency, combined with data-consistency alignment to match measured projections. Experiments on synthetic and real Cone Beam CT (CBCT) data show state-of-the-art performance, with ablations that confirm the synergistic effects of the proposed pipeline. Finally, we show that the same principles also strengthen fast denoising U-Nets, yielding near-diffusion quality at a fraction of the computational cost.

---

### A systematic review of generative AI usage for IT project management

Authors: Ionut Anghel, Tudor Cioara
Published: 2026-04-23
Categories: cs.SE, cs.AI
arXiv: https://arxiv.org/abs/2604.21958v1

This paper aims to synthesize current knowledge on generative AI in IT project management using the PRISMA methodology to provide researchers with a comprehensive perspective on techniques, applications, adoption trends, limitations, and integration across project management tools and process groups. The analysis reveals a clear dominance of OpenAI's GPT in the included studies but relying primarily on prompt engineering, suggesting that research in this area remains at an exploratory stage. Finally, it identifies and discusses three promising research directions for AI-enabled project management, including process group-specific AI agents, project role-based AI agents, and hybrid collaborative networks that enable human-guided orchestration.

---

### MambaCSP: Hybrid-Attention State Space Models for Hardware-Efficient Channel State Prediction

Authors: Aladin Djuhera, Haris Gacanin, Holger Boche
Published: 2026-04-23
Categories: cs.IT, cs.AI, cs.LG, eess.SP
arXiv: https://arxiv.org/abs/2604.21957v2

Recent works have demonstrated that attention-based transformer and large language model (LLM) architectures can achieve strong channel state prediction (CSP) performance by capturing long-range temporal dependencies across channel state information (CSI) sequences. However, these models suffer from quadratic scaling in sequence length, leading to substantial computational cost, memory consumption, and inference latency, which limits their applicability in real-time and resource-constrained wireless deployments. In this paper, we investigate whether selective state space models (SSMs) can serve as a hardware-efficient alternative for CSI prediction. We propose MambaCSP, a hybrid-attention SSM architecture that replaces LLM-based prediction backbones with a linear-time Mamba model. To overcome the local-only dependencies of pure SSMs, we introduce lightweight patch-mixer attention layers that periodically inject cross-token attentions, helping with long-context CSI prediction. Extensive MISO-OFDM simulations show that MambaCSP improves prediction accuracy over LLM-based approaches by 9-12%, while delivering up to 3.0x higher throughput, 2.6x lower VRAM usage, and 2.9x faster inference. Our results demonstrate that hybrid state space architectures provide a promising direction for scalable and hardware-efficient AI-native CSI prediction in future wireless networks.

---

### MambaCSP: Hybrid-Attention State Space Models for Hardware-Efficient Channel State Prediction

Authors: Aladin Djuhera, Haris Gacanin, Holger Boche
Published: 2026-04-23
Categories: cs.IT, cs.AI, cs.LG, eess.SP
arXiv: https://arxiv.org/abs/2604.21957v1

Recent works have demonstrated that attention-based transformer and large language model (LLM) architectures can achieve strong channel state prediction (CSP) performance by capturing long-range temporal dependencies across channel state information (CSI) sequences. However, these models suffer from quadratic scaling in sequence length, leading to substantial computational cost, memory consumption, and inference latency, which limits their applicability in real-time and resource-constrained wireless deployments. In this paper, we investigate whether selective state space models (SSMs) can serve as a hardware-efficient alternative for CSI prediction. We propose MambaCSP, a hybrid-attention SSM architecture that replaces LLM-based prediction backbones with a linear-time Mamba model. To overcome the local-only dependencies of pure SSMs, we introduce lightweight patch-mixer attention layers that periodically inject cross-token attentions, helping with long-context CSI prediction. Extensive MISO-OFDM simulations show that MambaCSP improves prediction accuracy over LLM-based approaches by 9-12%, while delivering up to 3.0x higher throughput, 2.6x lower VRAM usage, and 2.9x faster inference. Our results demonstrate that hybrid state space architectures provide a promising direction for scalable and hardware-efficient AI-native CSI prediction in future wireless networks.

---

### Conditional anomaly detection using soft harmonic functions: An application to clinical alerting

Authors: Michal Valko, Hamed Valizadegan, Branislav Kveton, Gregory F. Cooper, Milos Hauskrecht
Published: 2026-04-23
Categories: cs.LG
arXiv: https://arxiv.org/abs/2604.21956v1

Timely detection of concerning events is an important problem in clinical practice. In this paper, we consider the problem of conditional anomaly detection that aims to identify data instances with an unusual response, such as the omission of an important lab test. We develop a new non-parametric approach for conditional anomaly detection based on the soft harmonic solution, with which we estimate the confidence of the label to detect anomalous mislabeling. We further regularize the solution to avoid the detection of isolated examples and examples on the boundary of the distribution support. We demonstrate the efficacy of the proposed method in detecting unusual labels on a real-world electronic health record dataset and compare it to several baseline approaches.

---

### Performance Anomaly Detection in Athletics: A Benchmarking System with Visual Analytics

Authors: Blessed Madukoma, Prasenjit Mitra
Published: 2026-04-23
Categories: cs.LG, cs.CY
arXiv: https://arxiv.org/abs/2604.21953v1

Anti-doping programs rely on biological testing to detect performance-enhancing drugs, but such testing costs over $800 per sample and is limited by short detection windows for many prohibited substances. These constraints leave large portions of athletes without regular testing, motivating complementary screening approaches that analyze routine competition results to identify suspicious performance patterns. We present a system that processes 1.6 million athletics performances from over 19,000 competitions (2010-2025) using eight detection methods ranging from statistical rules to machine learning and trajectory analysis. We validate all methods against publicly confirmed anti-doping violations to measure their effectiveness in identifying sanctioned athletes. Trajectory-based methods, which compare performances to expected career progression, achieve the best balance between detecting violations and limiting false alarms, though all methods face challenges from incomplete data and rare confirmed violations. The system provides an interactive interface for expert-driven investigation, emphasizing transparency and human judgment to support, rather than replace, established anti-doping processes.

---

### Focus Session: Hardware and Software Techniques for Accelerating Multimodal Foundation Models

Authors: Muhammad Shafique, Abdul Basit, Muhammad Abdullah Hanif, Alberto Marchisio, Rachmad Vidya Wicaksana Putra, Minghao Shao
Published: 2026-04-23
Categories: cs.LG, cs.AI, cs.AR, cs.NE, cs.RO
arXiv: https://arxiv.org/abs/2604.21952v1

This work presents a multi-layered methodology for efficiently accelerating multimodal foundation models (MFMs). It combines hardware and software co-design of transformer blocks with an optimization pipeline that reduces computational and memory requirements. During model development, it employs performance enhancements through fine-tuning for domain-specific adaptation. Our methodology further incorporates hardware and software techniques for optimizing MFMs. Specifically, it employs MFM compression using hierarchy-aware mixed-precision quantization and structural pruning for transformer blocks and MLP channels. It also optimizes operations through speculative decoding, model cascading that routes queries through a small-to-large cascade and uses lightweight self-tests to determine when to escalate to larger models, as well as co-optimization of sequence length, visual resolution & stride, and graph-level operator fusion. To efficiently execute the model, the processing dataflow is optimized based on the underlying hardware architecture together with memory-efficient attention to meet on-chip bandwidth and latency budgets. To support this, a specialized hardware accelerator for the transformer workloads is employed, which can be developed through expert design or an LLM-aided design approach. We demonstrate the effectiveness of the proposed methodology on medical-MFMs and on code generation tasks, and conclude with extensions toward energy-efficient spiking-MFMs.

---

### Feedback Over Form: Why Execution Feedback Matters More Than Pipeline Topology in 1-3B Code Generation

Authors: Charles Junichi McAndrews
Published: 2026-04-23
Categories: cs.SE, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2604.21950v1

Small language models (1-3B) are practical to run locally, but individually limited on harder code generation tasks. We ask whether composing them into pipelines can recover some of that lost capability. We study code generation pipelines built from 1-3B models with execution feedback, and use a NEAT-inspired evolutionary search to test whether more complex pipeline structure helps beyond a simple refinement loop. We evaluate on HumanEval (164 problems) and sanitized MBPP (427 problems), all with local inference on a single laptop. Self-refinement with execution feedback improves code generation by more than 4 standard deviations on both benchmarks. The gains are narrow in mechanism: refinement fixes many runtime errors (especially NameError and SyntaxError), but rarely fixes logic errors such as AssertionError. Within our tested general-purpose model pool, generator identity mattered less than refiner capability: a 1.5B generator paired with a 3B refiner matched a 3B model doing both roles. Early stopping is essential; without it, every iteration is net-negative. The code-specialized models outperform every general-purpose pipeline configuration, suggesting model specialization matters more than pipeline architecture. Preliminary text-only pipeline experiments without execution feedback did not show gains at this scale. In our constrained search space, evolutionary search mostly rediscovered the same simple generate-execute-refine loop we found manually, with no clearly significant gain from added topology. Single-evaluation fitness inflates results by 5-7 percent, selecting lucky genomes over good ones. On these benchmarks at 1-3B scale, execution feedback mattered more than added pipeline complexity in determining whether composition helped.

---

### Seeing Fast and Slow: Learning the Flow of Time in Videos

Authors: Yen-Siang Wu, Rundong Luo, Jingsen Zhu, Tao Tu, Ali Farhadi, Matthew Wallingford, Yu-Chiang Frank Wang, Steve Marschner, Wei-Chiu Ma
Published: 2026-04-23
Categories: cs.CV, cs.AI, cs.GR
arXiv: https://arxiv.org/abs/2604.21931v1

How can we tell whether a video has been sped up or slowed down? How can we generate videos at different speeds? Although videos have been central to modern computer vision research, little attention has been paid to perceiving and controlling the passage of time. In this paper, we study time as a learnable visual concept and develop models for reasoning about and manipulating the flow of time in videos. We first exploit the multimodal cues and temporal structure naturally present in videos to learn, in a self-supervised manner, to detect speed changes and estimate playback speed. We then show that these learned temporal reasoning models enable us to curate the largest slow-motion video dataset to date from noisy in-the-wild sources. Such slow-motion footage, typically filmed by high-speed cameras, contains substantially richer temporal detail than standard videos. Using this data, we further develop models capable of temporal control, including speed-conditioned video generation, which produces motion at specified playback speed, and temporal super-resolution, which tranforms low-FPS, blurry videos into high-FPS sequences with fine-grained temporal details. Our findings highlight time as a manipulable, perceptual dimension in video learning, opening doors to temporally controllable video generation, temporal forensics detection, and potentially richer world-models that understand how events unfold over time.

---

### Temporal Taskification in Streaming Continual Learning: A Source of Evaluation Instability

Authors: Nicolae Filat, Ahmed Hussain, Konstantinos Kalogiannis, Elena Burceanu
Published: 2026-04-23
Categories: cs.LG
arXiv: https://arxiv.org/abs/2604.21930v1

Streaming Continual Learning (CL) typically converts a continuous stream into a sequence of discrete tasks through temporal partitioning. We argue that this temporal taskification step is not a neutral preprocessing choice, but a structural component of evaluation: different valid splits of the same stream can induce different CL regimes and therefore different benchmark conclusions. To study this effect, we introduce a taskification-level framework based on plasticity and stability profiles, a profile distance between taskifications, and Boundary-Profile Sensitivity (BPS), which diagnoses how strongly small boundary perturbations alter the induced regime before any CL model is trained. We evaluate continual finetuning, Experience Replay, Elastic Weight Consolidation, and Learning without Forgetting on network traffic forecasting with CESNET-Timeseries24, keeping the stream, model, and training budget fixed while varying only the temporal taskification. Across 9-, 30-, and 44-day splits, we observe substantial changes in forecasting error, forgetting, and backward transfer, showing that taskification alone can materially affect CL evaluation. We further find that shorter taskifications induce noisier distribution-level patterns, larger structural distances, and higher BPS, indicating greater sensitivity to boundary perturbations. These results show that benchmark conclusions in streaming CL depend not only on the learner and the data stream, but also on how that stream is taskified, motivating temporal taskification as a first-class evaluation variable.

---

### Evaluation of Automatic Speech Recognition Using Generative Large Language Models

Authors: Thibault Bañeras-Roux, Shashi Kumar, Driss Khalil, Sergio Burdisso, Petr Motlicek, Shiran Liu, Mickael Rouvier, Jane Wottawa, Richard Dufour
Published: 2026-04-23
Categories: cs.CL
arXiv: https://arxiv.org/abs/2604.21928v3

Automatic Speech Recognition (ASR) is traditionally evaluated using Word Error Rate (WER), a metric that is insensitive to meaning. Embedding-based semantic metrics are better correlated with human perception, but decoder-based Large Language Models (LLMs) remain underexplored for this task. This paper evaluates their relevance through three approaches: (1) selecting the best hypothesis between two candidates, (2) computing semantic distance using generative embeddings, and (3) qualitative classification of errors. On the HATS dataset, the best LLMs achieve 92--94\% agreement with human annotators for hypothesis selection, compared to 63\% for WER, also outperforming semantic metrics. Embeddings from decoder-based LLMs show performance comparable to encoder models. Finally, LLMs offer a promising direction for interpretable and semantic ASR evaluation.

---

### Fine-Tuning Regimes Define Distinct Continual Learning Problems

Authors: Paul-Tiberiu Iordache, Elena Burceanu
Published: 2026-04-23
Categories: cs.LG
arXiv: https://arxiv.org/abs/2604.21927v3

Continual learning (CL) studies how models acquire tasks sequentially while retaining previously learned knowledge. Despite substantial progress in benchmarking CL methods, comparative evaluations typically keep the fine-tuning regime fixed. In this paper, we argue that the fine-tuning regime, defined by the trainable parameter subspace, is itself a key evaluation variable. We formalize adaptation regimes as projected optimization over fixed trainable subspaces, showing that changing the trainable depth alters the effective update signal through which both current task fitting and knowledge preservation operate. This analysis motivates the hypothesis that method comparisons need not be invariant across regimes. We test this hypothesis in task incremental CL, five trainable depth regimes, and four standard methods: online EWC, LwF, SI, and GEM. Across five benchmark datasets, namely MNIST, Fashion MNIST, KMNIST, QMNIST, and CIFAR-100, and across 11 task orders per dataset, we find that the relative ranking of methods is not consistently preserved across regimes. We further show that deeper adaptation regimes are associated with larger update magnitudes, higher forgetting, and a stronger relationship between the two. These results show that comparative conclusions in CL can depend strongly on the chosen fine-tuning regime, motivating regime-aware evaluation protocols that treat trainable depth as an explicit experimental factor.

---

### Seeing Without Eyes: 4D Human-Scene Understanding from Wearable IMUs

Authors: Hao-Yu Hsu, Tianhang Cheng, Jing Wen, Alexander G. Schwing, Shenlong Wang
Published: 2026-04-23
Categories: cs.CV
arXiv: https://arxiv.org/abs/2604.21926v1

Understanding human activities and their surrounding environments typically relies on visual perception, yet cameras pose persistent challenges in privacy, safety, energy efficiency, and scalability. We explore an alternative: 4D perception without vision. Its goal is to reconstruct human motion and 3D scene layouts purely from everyday wearable sensors. For this we introduce IMU-to-4D, a framework that repurposes large language models for non-visual spatiotemporal understanding of human-scene dynamics. IMU-to-4D uses data from a few inertial sensors from earbuds, watches, or smartphones and predicts detailed 4D human motion together with coarse scene structure. Experiments across diverse human-scene datasets show that IMU-to-4D yields more coherent and temporally stable results than SoTA cascaded pipelines, suggesting wearable motion sensors alone can support rich 4D understanding.

---

### The Sample Complexity of Multicalibration

Authors: Natalie Collina, Jiuyao Lu, Georgy Noarov, Aaron Roth
Published: 2026-04-23
Categories: cs.LG, math.ST, stat.ML
arXiv: https://arxiv.org/abs/2604.21923v1

We study the minimax sample complexity of multicalibration in the batch setting. A learner observes $n$ i.i.d. samples from an unknown distribution and must output a (possibly randomized) predictor whose population multicalibration error, measured by Expected Calibration Error (ECE), is at most $\varepsilon$ with respect to a given family of groups. For every fixed $κ> 0$, in the regime $|G|\le \varepsilon^{-κ}$, we prove that $\widetildeΘ(\varepsilon^{-3})$ samples are necessary and sufficient, up to polylogarithmic factors. The lower bound holds even for randomized predictors, and the upper bound is realized by a randomized predictor obtained via an online-to-batch reduction. This separates the sample complexity of multicalibration from that of marginal calibration, which scales as $\widetildeΘ(\varepsilon^{-2})$, and shows that mean-ECE multicalibration is as difficult in the batch setting as it is in the online setting, in contrast to marginal calibration which is strictly more difficult in the online setting. In contrast we observe that for $κ= 0$, the sample complexity of multicalibration remains $\widetildeΘ(\varepsilon^{-2})$ exhibiting a sharp threshold phenomenon. More generally, we establish matching upper and lower bounds, up to polylogarithmic factors, for a weighted $L_p$ multicalibration metric for all $1 \le p \le 2$, with optimal exponent $3/p$. We also extend the lower-bound template to a regular class of elicitable properties, and combine it with the online upper bounds of Hu et al. (2025) to obtain matching bounds for calibrating properties including expectiles and bounded-density quantiles.

---

### Context Unrolling in Omni Models

Authors: Ceyuan Yang, Zhijie Lin, Yang Zhao, Fei Xiao, Hao He, Qi Zhao, Chaorui Deng, Kunchang Li, Zihan Ding, Yuwei Guo, Fuyun Wang, Fangqi Zhu et al.
Published: 2026-04-23
Categories: cs.CV
arXiv: https://arxiv.org/abs/2604.21921v1

We present Omni, a unified multimodal model natively trained on diverse modalities, including text, images, videos, 3D geometry, and hidden representations. We find that such training enables Context Unrolling, where the model explicitly reasons across multiple modal representations before producing predictions. This process enables the model to aggregate complementary information across heterogeneous modalities, facilitating a more faithful approximation of the shared multimodal knowledge manifold and improving downstream reasoning fidelity. As a result, Omni achieves strong performance on both multimodal generation and understanding benchmarks, while demonstrating advanced multimodal reasoning capabilities, including in-context generation of text, image, video, and 3D geometry.

---

### MathDuels: A Self-Play Benchmark That Grows

Authors: Zhiqiu Xu, Shibo Jin, Shreya Arya, Mayur Naik
Published: 2026-04-23
Categories: cs.CL, cs.SE
arXiv: https://arxiv.org/abs/2604.21916v3

As frontier language models attain near-ceiling performance on static mathematical benchmarks, existing evaluations are increasingly unable to differentiate model capabilities, largely because they cast models solely as solvers of fixed problem sets. We introduce MathDuels, a self-play benchmark in which models occupy dual roles: each authors math problems under adversarial prompting and solves problems authored by every other participant. Problems are produced through a three-stage generation pipeline (meta-prompting, problem generation, and difficulty amplification), and validated by an independent verifier that excludes ill-posed questions. A Rasch model (Rasch, 1993) jointly estimates solver abilities and problem difficulties; author quality is derived from the difficulties of each model's authored problems. Experiments across 19 frontier models reveal that authoring and solving capabilities are partially decoupled, and that dual-role evaluation reveals capability separations invisible in single-role benchmarks. As newer models enter the arena, they produce problems that defeat previously dominant solvers, so the benchmark's difficulty co-evolves with participant strength rather than saturating at a fixed ceiling. We host a public leaderboard that updates as new models are released.

---

### MathDuels: Evaluating LLMs as Problem Posers and Solvers

Authors: Zhiqiu Xu, Shibo Jin, Shreya Arya, Mayur Naik
Published: 2026-04-23
Categories: cs.CL, cs.SE
arXiv: https://arxiv.org/abs/2604.21916v2

As frontier language models attain near-ceiling performance on static mathematical benchmarks, existing evaluations are increasingly unable to differentiate model capabilities, largely because they cast models solely as solvers of fixed problem sets. We introduce MathDuels, a self-play benchmark in which models occupy dual roles: each authors math problems under adversarial prompting and solves problems authored by every other participant. Problems are produced through a three-stage generation pipeline (meta-prompting, problem generation, and difficulty amplification), and validated by an independent verifier that excludes ill-posed questions. A Rasch model (Rasch, 1993) jointly estimates solver abilities and problem difficulties; author quality is derived from the difficulties of each model's authored problems. Experiments across 19 frontier models reveal that authoring and solving capabilities are partially decoupled, and that dual-role evaluation reveals capability separations invisible in single-role benchmarks. As newer models enter the arena, they produce problems that defeat previously dominant solvers, so the benchmark's difficulty co-evolves with participant strength rather than saturating at a fixed ceiling. We host a public leaderboard that updates as new models are released.

---

### Vista4D: Video Reshooting with 4D Point Clouds

Authors: Kuan Heng Lin, Zhizheng Liu, Pablo Salamanca, Yash Kant, Ryan Burgert, Yuancheng Xu, Koichi Namekata, Yiwei Zhao, Bolei Zhou, Micah Goldblum, Paul Debevec, Ning Yu
Published: 2026-04-23
Categories: cs.CV
arXiv: https://arxiv.org/abs/2604.21915v1

We present Vista4D, a robust and flexible video reshooting framework that grounds the input video and target cameras in a 4D point cloud. Specifically, given an input video, our method re-synthesizes the scene with the same dynamics from a different camera trajectory and viewpoint. Existing video reshooting methods often struggle with depth estimation artifacts of real-world dynamic videos, while also failing to preserve content appearance and failing to maintain precise camera control for challenging new trajectories. We build a 4D-grounded point cloud representation with static pixel segmentation and 4D reconstruction to explicitly preserve seen content and provide rich camera signals, and we train with reconstructed multiview dynamic data for robustness against point cloud artifacts during real-world inference. Our results demonstrate improved 4D consistency, camera control, and visual quality compared to state-of-the-art baselines under a variety of videos and camera paths. Moreover, our method generalizes to real-world applications such as dynamic scene expansion and 4D scene recomposition. See our project page for results, code, and models: https://eyeline-labs.github.io/Vista4D

---

### When Prompts Override Vision: Prompt-Induced Hallucinations in LVLMs

Authors: Pegah Khayatan, Jayneel Parekh, Arnaud Dapogny, Mustafa Shukor, Alasdair Newson, Matthieu Cord
Published: 2026-04-23
Categories: cs.CV, cs.AI, cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2604.21911v1

Despite impressive progress in capabilities of large vision-language models (LVLMs), these systems remain vulnerable to hallucinations, i.e., outputs that are not grounded in the visual input. Prior work has attributed hallucinations in LVLMs to factors such as limitations of the vision backbone or the dominance of the language component, yet the relative importance of these factors remains unclear. To resolve this ambiguity, We propose HalluScope, a benchmark to better understand the extent to which different factors induce hallucinations. Our analysis indicates that hallucinations largely stem from excessive reliance on textual priors and background knowledge, especially information introduced through textual instructions. To mitigate hallucinations induced by textual instruction priors, we propose HalluVL-DPO, a framework for fine-tuning off-the-shelf LVLMs towards more visually grounded responses. HalluVL-DPO leverages preference optimization using a curated training dataset that we construct, guiding the model to prefer grounded responses over hallucinated ones. We demonstrate that our optimized model effectively mitigates the targeted hallucination failure mode, while preserving or improving performance on other hallucination benchmarks and visual capability evaluations. To support reproducibility and further research, we will publicly release our evaluation benchmark, preference training dataset, and code at https://pegah-kh.github.io/projects/prompts-override-vision/ .

---

### From Research Question to Scientific Workflow: Leveraging Agentic AI for Science Automation

Authors: Bartosz Balis, Michal Orzechowski, Piotr Kica, Michal Dygas, Michal Kuszewski
Published: 2026-04-23
Categories: cs.AI
arXiv: https://arxiv.org/abs/2604.21910v1

Scientific workflow systems automate execution -- scheduling, fault tolerance, resource management -- but not the semantic translation that precedes it. Scientists still manually convert research questions into workflow specifications, a task requiring both domain knowledge and infrastructure expertise. We propose an agentic architecture that closes this gap through three layers: an LLM interprets natural language into structured intents (semantic layer); validated generators produce reproducible workflow DAGs (deterministic layer); and domain experts author ``Skills'': markdown documents encoding vocabulary mappings, parameter constraints, and optimization strategies (knowledge layer). This decomposition confines LLM non-determinism to intent extraction: identical intents always yield identical workflows. We implement and evaluate the architecture on the 1000 Genomes population genetics workflow and Hyperflow WMS running on Kubernetes. In an ablation study on 150 queries, Skills raise full-match intent accuracy from 44% to 83%; skill-driven deferred workflow generation reduces data transfer by 92\%; and the end-to-end pipeline completes queries on Kubernetes with LLM overhead below 15 seconds and cost under $0.001 per query.

---

### Directional Confusions Reveal Divergent Inductive Biases Through Rate-Distortion Geometry in Human and Machine Vision

Authors: Leyla Roksan Caglar, Pedro A. M. Mediano, Baihan Lin
Published: 2026-04-23
Categories: cs.CV, cs.IT, q-bio.NC
arXiv: https://arxiv.org/abs/2604.21909v2

To humans, a robin seems more like a bird than a bird seems like a robin, but does this asymmetry also hold for machine vision? Humans and modern vision models can match each other in accuracy while making systematically different kinds of errors, differing not in how often they fail, but in who gets mistaken for whom. We show these directional confusions reveal distinct inductive biases invisible to accuracy alone. Using matched human and deep neural network responses on a natural-image categorization task under 12 perturbation types, we quantify asymmetry in confusion matrices and link its organization to the geometry of the information--error trade-off - how efficiently, and how gracefully, a system generalizes under distortion. We find that humans exhibit broad but weak asymmetries across many class pairs, whereas deep vision models show sparser, stronger directional collapses into a few dominant categories. Robustness training reduces overall asymmetry magnitude but fails to recover this human-like distributed structure. Generative simulations further show that these two asymmetry organizations shift the trade-off geometry in opposite directions even at matched accuracy, explaining why the same scalar asymmetry score can reflect fundamentally different generalization strategies. Together, these results establish directional confusion structure as a sensitive, interpretable signature of inductive bias that accuracy-based evaluation cannot recover.

---

### Low-Rank Adaptation Redux for Large Models

Authors: Bingcong Li, Yilang Zhang, Georgios B. Giannakis
Published: 2026-04-23
Categories: cs.LG, eess.SP
arXiv: https://arxiv.org/abs/2604.21905v1

Low-rank adaptation (LoRA) has emerged as the de facto standard for parameter-efficient fine-tuning (PEFT) of foundation models, enabling the adaptation of billion-parameter networks with minimal computational and memory overhead. Despite its empirical success and rapid proliferation of variants, it remains elusive which architectural choices, optimization techniques, and deployment constraints should guide practical method selection. This overview revisits LoRA through the lens of signal processing (SP), bridging modern adapter designs with classical low-rank modeling tools and inverse problems, as well as highlighting how SP principles can inform principled advances of fine-tuning approaches. Rather than providing a comprehensive enumeration and empirical comparisons of LoRA variants, emphasis is placed on the technical mechanisms underpinning these approaches to justify their effectiveness. These advances are categorized into three complementary axes: architectural design, efficient optimization, and pertinent applications. The first axis builds on singular value decomposition (SVD)-based factorization, rank-augmentation constructions, and cross-layer tensorization, while the second axis deals with initialization, alternating solvers, gauge-invariant optimization, and parameterization-aware methods. Beyond fine-tuning, emerging applications of LoRA are accounted across the entire lifecycle of large models, ranging from pre- and post-training to serving/deployment. Finally, open research directions are outlined at the confluence of SP and deep learning to catalyze a bidirectional frontier: classical SP tools provide a principled vocabulary for designing principled PEFT methods, while the unique challenges facing modern deep learning, especially the overwhelming scale and prohibitive overhead, also offer new research lines benefiting the SP community in return.

---

### UniGenDet: A Unified Generative-Discriminative Framework for Co-Evolutionary Image Generation and Generated Image Detection

Authors: Yanran Zhang, Wenzhao Zheng, Yifei Li, Bingyao Yu, Yu Zheng, Lei Chen, Jiwen Lu, Jie Zhou
Published: 2026-04-23
Categories: cs.CV
arXiv: https://arxiv.org/abs/2604.21904v1

In recent years, significant progress has been made in both image generation and generated image detection. Despite their rapid, yet largely independent, development, these two fields have evolved distinct architectural paradigms: the former predominantly relies on generative networks, while the latter favors discriminative frameworks. A recent trend in both domains is the use of adversarial information to enhance performance, revealing potential for synergy. However, the significant architectural divergence between them presents considerable challenges. Departing from previous approaches, we propose UniGenDet: a Unified generative-discriminative framework for co-evolutionary image Generation and generated image Detection. To bridge the task gap, we design a symbiotic multimodal self-attention mechanism and a unified fine-tuning algorithm. This synergy allows the generation task to improve the interpretability of authenticity identification, while authenticity criteria guide the creation of higher-fidelity images. Furthermore, we introduce a detector-informed generative alignment mechanism to facilitate seamless information exchange. Extensive experiments on multiple datasets demonstrate that our method achieves state-of-the-art performance. Code: \href{https://github.com/Zhangyr2022/UniGenDet}{https://github.com/Zhangyr2022/UniGenDet}.

---

### A Scale-Adaptive Framework for Joint Spatiotemporal Super-Resolution with Diffusion Models

Authors: Max Defez, Filippo Quarenghi, Mathieu Vrac, Stephan Mandt, Tom Beucler
Published: 2026-04-23
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2604.21903v1

Deep-learning video super-resolution has progressed rapidly, but climate applications typically super-resolve (increase resolution) either space or time, and joint spatiotemporal models are often designed for a single pair of super-resolution (SR) factors (upscaling spatial and temporal ratio between the low-resolution sequence and the high-resolution sequence), limiting transfer across spatial resolutions and temporal cadences (frame rates). We present a scale-adaptive framework that reuses the same architecture across factors by decomposing spatiotemporal SR into a deterministic prediction of the conditional mean, with attention, and a residual conditional diffusion model, with an optional mass-conservation (same precipitation amount in inputs and outputs) transform to preserve aggregated totals. Assuming that larger SR factors primarily increase underdetermination (hence required context and residual uncertainty) rather than changing the conditional-mean structure, scale adaptivity is achieved by retuning three factor-dependent hyperparameters before retraining: the diffusion noise schedule amplitude beta (larger for larger factors to increase diversity), the temporal context length L (set to maintain comparable attention horizons across cadences) and optionally a third, the mass-conservation function f (tapered to limit the amplification of extremes for large factors). Demonstrated on reanalysis precipitation over France (Comephore), the same architecture spans super-resolution factors from 1 to 25 in space and 1 to 6 in time, yielding a reusable architecture and tuning recipe for joint spatiotemporal super-resolution across scales.

---

### GiVA: Gradient-Informed Bases for Vector-Based Adaptation

Authors: Neeraj Gangwar, Rishabh Deshmukh, Michael Shavlovsky, Hancao Li, Vivek Mittal, Lexing Ying, Nickvash Kani
Published: 2026-04-23
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2604.21901v1

As model sizes continue to grow, parameter-efficient fine-tuning has emerged as a powerful alternative to full fine-tuning. While LoRA is widely adopted among these methods, recent research has explored vector-based adaptation methods due to their extreme parameter efficiency. However, these methods typically require substantially higher ranks than LoRA to match its performance, leading to increased training costs. This work introduces GiVA, a gradient-based initialization strategy for vector-based adaptation. It achieves training times comparable to LoRA and maintains the extreme parameter efficiency of vector-based adaptation. We evaluate GiVA across diverse benchmarks, including natural language understanding, natural language generation, and image classification. Experiments show that our approach consistently outperforms or achieves performance competitive with existing vector-based adaptation methods and LoRA while reducing rank requirements by a factor of eight ($8\times$).

---