# Research Papers — 2026-05 · bundle page 10 of 326 > Full-text markdown of papers #451–500 of 16266 published in 2026-05, newest first. AI-readable mirrors of the full 16266-paper 2026-05 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-05/page/10 [← Previous bundle](https://aipod.dev/papers/archive/2026-05/page/9.md) [Next bundle →](https://aipod.dev/papers/archive/2026-05/page/11.md) ### Quality-Diversity Evolution for Discovering Diverse Vulnerabilities in LLM Safety Authors: Subhadip Mitra Published: 2026-05-30 Categories: cs.CR, cs.CL, cs.ET, cs.LG, cs.NE arXiv: https://arxiv.org/abs/2606.00801v1 Current approaches to LLM adversarial testing suffer from coverage gaps: manual red-teaming does not scale, LLM-as-attacker methods exhibit mode collapse, and gradient-based approaches produce uninterpretable gibberish. We introduce a quality-diversity evolutionary framework that operates at the semantic level, evolving interpretable attack strategies rather than token sequences. Using MAP-Elites, we maintain a diverse archive of attacks across behavioral dimensions (strategy type, encoding method, prompt length). In experiments across GPT-4o-mini, Claude 3.5 Sonnet, Gemini 2.0 Flash, and an open-weight coding model (Devstral-small-2), we discover distinct vulnerability profiles: GPT-4o-mini is vulnerable to hypothetical and multi-turn framing combined with ROT13 encoding (fitness 0.8), Gemini to direct attacks with ROT13 and multi-turn with Leetspeak (0.8), while Claude shows uniformly ambiguous responses across all strategies (max 0.4). The semantic representation produces interpretable attacks that reveal systematic, model-specific weaknesses, providing actionable insights for improving LLM safety and a reproducible baseline for evaluating future frontier models. Code and experiment artifacts are released at https://github.com/bassrehab/red-queen. --- ### DASH: Dual-Branch Score Distillation for Guidance-Calibrated Compact Diffusion Models Authors: Abdullah Al Shafi, Kazi Saeed Alam, Sk Imran Hossain, Engelbert Mephu Nguifo Published: 2026-05-30 Categories: cs.CV, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.00798v2 Parameter compression of class-conditional diffusion models exposes a structural limitation in output-level distillation: supervising only the guided output leaves the two score branches non-identifiable, so the classifier-free guidance gap is not determined in the student. The zero-loss set admits degenerate solutions in which both branches collapse toward identical predictions, and guidance loses its effect at inference despite low training loss. For each fixed input the objective is exactly flat in a branch-output direction at every residual value, so the ambiguity is a property of the output-level objective rather than a poor local minimum. This paper introduces DASH, which supervises the conditional and unconditional branches independently. An anchor term regularises the conditional prediction toward ground-truth noise, and the teacher's final learned per-timestep curriculum transfers into the student as a frozen prior. Across CIFAR-10, CIFAR-100, and class-conditional ImageNet-64, a more than 5x compressed student stays within four FID points of its teacher and recovers at least 89% of the teacher's guidance-gap magnitude, where no other two-branch baseline with defined calibration exceeds 82%. Ablation isolates unconditional supervision as the term that separates this formulation from output-level distillation. The account is quantitative: the null space fixes the ratio between a composite student's two branch errors, and the measured ratios approximately match that prediction on all three datasets. --- ### DASH: Dual-Branch Score Distillation for Guidance-Calibrated Compact Diffusion Models Authors: Abdullah Al Shafi, Kazi Saeed Alam, Sk Imran Hossain, Engelbert Mephu Nguifo Published: 2026-05-30 Categories: cs.CV, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.00798v1 Parameter compression of class-conditional diffusion models reveals an underexplored limitation in output-level distillation: the unconditional score branch remains unsupervised, leaving the classifier-free guidance gap underdetermined in the student. This gap, amplified at every denoising step, admits degenerate solutions where both branches collapse toward identical predictions, rendering guidance ineffective despite low output-level training loss. This paper introduces DASH, a dual-branch distillation framework that independently supervises both score branches, uniquely specifying target branch outputs for each training sample through independent branch constraints, with an anchor term regularising conditional predictions toward ground-truth noise. The framework further introduces TIRT Transfer, which copies the teacher's converged per-timestep importance curriculum into the student as a frozen prior, eliminating the need to relearn it within limited distillation budgets. Experiments on CIFAR-10 and CIFAR-100 demonstrate that 5.9x compression maintains quality within 4 FID points of the teacher at 50-step DDIM sampling, considerably outperforming training from scratch with guidance fidelity well preserved. Ablation studies confirm that unconditional supervision is the dominant contribution, accounting for over 60% of total distillation gain. Curriculum transfer and anchor regularisation provide complementary benefit, together validating dual-branch constraints as empirically essential for guidance-preserving compression. --- ### Robust inference for risk heterogeneity under group imbalance Authors: Mengqi Xu, Subha Maity, Joel Dubin Published: 2026-05-30 Categories: stat.ME, stat.AP, stat.ML arXiv: https://arxiv.org/abs/2606.00797v1 Population-level heterogeneity is ubiquitous in biomedical data, where differences across demographic or clinical subgroups can substantially alter risk patterns. For example, in intensive care unit (ICU) studies, the mortality risk associated with specific admission diagnoses can vary across ethnic groups. Existing approaches for detecting risk heterogeneity are often sensitive to baseline model misspecification and regularization bias, both of which commonly arise in practice. In this paper, we propose a robust framework for inferring risk heterogeneity between two populations using Neyman orthogonality, which yields estimators that are locally insensitive to nuisance parameter estimation error. The proposed estimator is consistent and asymptotically normal, and simulation studies demonstrate that in finite samples our method substantially reduces bias and improves inferential stability compared with standard likelihood-based approaches. In an application to the eICU Collaborative Research Database, our method reveals clinically meaningful ethnicity-specific heterogeneity in admission diagnoses for in-hospital mortality that standard likelihood-based methods fail to detect. --- ### Extending Causal Metamodeling to a non-Markovian Queue Authors: Pracheta Amaranath, Anant Bhide, David Jensen, Peter Haas Published: 2026-05-30 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2606.00795v1 Metamodels for discrete-event simulations approximate the behavior of simulation models without running expensive simulations. Prior work introduced modular dynamic Bayesian networks (MDBNs) -- a class of metamodels that can estimate a range of probabilistic and causal queries (PCQs) using a single, trained model -- but the method was limited to Markovian systems. In this paper, we initiate an extension of MDBNs to non-Markovian queues by approximating non-exponential distributions using phase-type distributions. This approach raises novel challenges, including balancing metamodeling accuracy and tractability when choosing the number of phases, efficiently learning metamodel parameters, and choosing the sampling interval that is used to approximate a continuous-time simulation by a discrete-time MDBN. We provide preliminary solutions to these challenges, yielding the first causal metamodeling technique for non-Markovian systems. Experiments on a G/M/1 queue demonstrate that the MDBN can produce accurate answers to PCQs with orders-of-magnitude speedup of inference times relative to direct simulation. --- ### Benchmark Dataset for Catalysis on 2D MXenes Authors: Pavlo Melnyk, Anmar Karmush, Mårten Wadenbäck, Ania Beatriz Rodríguez-Barrera, Johanna Rosen, Michael Felsberg, Jonas Björk Published: 2026-05-30 Categories: cond-mat.mtrl-sci, cs.LG arXiv: https://arxiv.org/abs/2606.00794v1 Merging first-principles calculations with machine learning (ML), we aim to accelerate the exploration of catalytic behaviour in novel materials. We focus on two-dimensional (2D) Ti$_2$CT$_y$ MXenes, whose versatile surface chemistry makes them particularly compelling candidates for catalysis. Resolving their composition and structure under realistic conditions exceeds the reach of standard density functional theory (DFT) due to computational cost. To address this challenge, we generate a comprehensive dataset of 50,000 DFT calculations for training and 10,000 for testing, encompassing both Ti$_2$CT$_y$ MXene configurations and molecular systems, along with an additional test dataset with 1000 genuinely new, larger systems to investigate how well models generalise. We train and validate widely used and competitive machine learning interatomic potential (MLIP) models, including EquiformerV2, MACE, MatRIS, and UPET, that accurately predict atomic forces and formation energies -- quantities that DFT must repeatedly compute for structural and catalytic investigations -- for these 2D materials. This combined DFT-ML framework achieves computational acceleration on the order of approximately $1-4 \cdot 10^3$ (on a CPU) while maintaining desired-level accuracy (approximately +/- $10$ meV/A for forces and approximately +/- $1$ meV for per-atom energies), paving the way for more efficient investigations of MXene catalytic behaviour. Moreover, we perform an extensive qualitative evaluation of the trained models, showcasing the importance of comprehensive simulation-based comparison beyond benchmark metrics. The dataset and the trained models with the code are available at https://huggingface.co/datasets/CatalystAnonymous/catalyst_mxenes. --- ### MBench: A Comprehensive Benchmark on Memory Capability for Video World Models Authors: Shengjun Zhang, Zhang Zhang, Simin Huang, Zhenyu Tang, Hanyang Wang, Chensheng Dai, Min Chen, Yifan Li, Yuxin Li, Yingjie Chen, Hao Liu, Chen Li et al. Published: 2026-05-30 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.00793v2 Recent advancements in video-based world models have demonstrated an unprecedented ability to synthesize high-fidelity visual sequences. However, a fundamental gap persists between visually plausible video generation and the functional requirements of a world model, particularly in maintaining a stable and reasonable internal state over extended temporal horizons. While existing benchmarks primarily emphasize visual quality, motion coherence, and text-video alignment, they largely overlook memory, the core capability of a world model to preserve consistency across long-term horizons and complex interactions. To address this gap, we present \textbf{MBench}, a comprehensive benchmark dedicated to quantifying and evaluating the memory capability of video world models. We systematically decompose the memory capability of video world models into three hierarchical and complementary core dimensions: entity consistency, environment consistency, and causal consistency, which are further refined into 12 quantifiable sub-dimensions for comprehensive characterization of long-term memory. Our benchmark is built upon rigorously curated real-captured long videos, and evaluated by rule-based quantitative matrices and VLM to enable objective and comprehensive consistency assessment. Extensive evaluations of mainstream state-of-the-art video world models reveal critical systemic limitations of existing methods in long-term state retention, providing a standardized benchmark and clear research direction to advance the field. --- ### DINO-GFSA: Geo-Localization via Semantic Gated Fusion and Mamba-based Sequential Aggregation Authors: Beier Hu, Yuanshen Guo, Jialu Cai, Chengwei Li, Yong Wang, Shunan Wu, Zhigang Wu Published: 2026-05-30 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.00784v1 Cross-view geo-localization (CVGL) is critical for Unmanned Aerial Vehicle (UAV) self-positioning and target localization in GNSS-denied environments. However, acquiring robust semantics while preserving finegrained spatial details remains challenging. To address this, we propose DINO-GFSA, a framework leveraging a LoRA (Low-Rank Adaptation) adapted DINOv3 (ViTL) backbone for parameter-efficient, high-capacity representation. Crucially, we introduce a Semantic Gated Residual Fusion module, which utilizes high-level semantics to selectively calibrate and integrate low-level spatial cues, effectively bridging the semantic gap. Furthermore, a Mamba-based Sequential Aggregation Head is designed to capture long-range spatial dependencies with linear complexity. Experiments demonstrate state-of-the-art performance on University-1652 and DenseUAV benchmarks, notably surpassing the previous best on DenseUAV by 3.48% on Recall@1. These results validate DINO-GFSA as a generalized, robust solution for UAV CVGL. --- ### Bayesian Inference of Nonlinear Malaria Dynamics in Ghana via an Ensemble Markov Chain Monte Carlo Sampler Authors: T. Ansah-Narh, Y. Asare Afrane, J. Bremang Tandoh Published: 2026-05-30 Categories: stat.AP, cs.AI, math.PR, stat.CO arXiv: https://arxiv.org/abs/2606.00783v1 Reliable quantification of malaria dynamics in sub-Saharan Africa is hindered by short, noisy, and spatially heterogeneous surveillance records. In Ghana, health-facility data from 2014 to 2023 reveal non-linear and age-specific fluctuations in hospital admissions, yet existing approaches struggle to capture stochastic variability or provide credible uncertainty bounds. This study develops a Bayesian nonlinear inference framework that integrates a cubic baseline with a damped oscillatory kernel, estimated via an affine-invariant ensemble Markov Chain Monte Carlo sampler. The framework accommodates limited data, models parameter uncertainty, and generates probabilistic forecasts for children under five years and individuals aged five years or more. Results show strong empirical adequacy ($R^2 = 0.9958$ for $<5$ years; $R^2 = 0.9956$ for $\geq 5$ years) with residual errors below $2\%$ and well-mixed posteriors confirming convergence. District-level analysis reveals pronounced spatial heterogeneity, with coefficients of variation ranging from $<0.07$ in urban centres such as Kumasi to $>3.3$ in peripheral districts such as Mpohor and Bia East. Forecasts for 2024-2026 indicate a gradual resurgence: from 137,000 to 149,000 cases among children under five years and from 348,000 to 375,000 cases among older individuals, with uncertainty widening over time. By producing probabilistic forecasts, this Bayesian framework provides a principled tool for anticipating malaria fluctuations and strengthening data-driven decision-making in Ghana's national malaria control strategy. --- ### FlowOVD: Learning Generative Latent Flows for Zero-shot Open-vocabulary Detection Authors: Yao Wei, Andrea Cavallaro, Changjae Oh Published: 2026-05-30 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.00782v1 Open-vocabulary object detection (OVD) has achieved remarkable progress through large-scale vision-language pre-training. Existing methods, however, typically formulate OVD as a discriminative prediction problem, where decoder queries are either static or initialized from encoder features, thus limiting their diversity and flexibility. In this paper, we introduce a generative perspective by modeling decoder query generation as a continuous transport process in latent space. We propose FlowOVD, a text-conditioned query generation framework based on rectified flow that progressively transforms text-agnostic queries into text-guided queries. By introducing continuous latent query dynamics into a vision-language model (VLM) based detector, our method avoids heuristic discrete query construction and enables more expressive semantic alignment for open-vocabulary detection. Without requiring additional training data, FlowOVD achieves 49.5 AP on COCO and 31.5 AP on LVIS, outperforming GroundingDINO by +1.2 AP (+2.5 %) and +4.1 AP (+15.0 %), respectively. The larger gain on the challenging long-tailed LVIS benchmark further highlights the effectiveness of continuous query generation for open-vocabulary generalization. --- ### Behavior-Invariant Task Representation Learning with Transformer-based World Models for Offline Meta-Reinforcement Learning Authors: Fuyuan Qian, Menglong Zhang, Song Wang, Quanying Liu Published: 2026-05-30 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2606.00780v1 Offline meta-reinforcement learning leverages static datasets to enable agents to generalize to unseen environments by combining offline efficiency with meta-learning adaptability, yet it faces key challenges from context and policy distribution shifts. These issues hinder agents from adapting to online environments, and are further exacerbated under sparse-reward settings. As a result, agents often become trapped in an inherent pattern dilemma, failing to achieve robust generalization. In this work, we propose a novel framework that integrates information-theoretic task representation learning with a Transformer-based stochastic world model. Our approach extracts task-defining latent variables that are invariant to behavior policy, thereby effectively mitigating the context distribution shift. To further handle policy shift and model exploitation, we apply a conservative value penalty to imagination-based rollouts, preventing the policy from exploiting model inaccuracies while maintaining robust adaptation. Extensive evaluations demonstrate that our method outperforms state-of-the-art approaches, with superior stability and generalization under out-of-distribution and sparse-reward settings. --- ### Latent Diffusion Pretraining for Crystal Property Prediction Authors: Shrimon Mukherjee, Kishalay Das, Partha Basuchowdhuri, Pawan Goyal, Niloy Ganguly Published: 2026-05-30 Categories: cs.LG arXiv: https://arxiv.org/abs/2606.00776v1 Fast and accurate prediction of crystal properties is a central challenge in new materials design. Graph neural networks and Transformer-based models have emerged as powerful tools for this task due to their ability to encode the local structural environment of atoms within a crystal. However, these models are data-hungry, and in practice, labeled data for crystal properties are scarce. Pretraining-finetuning strategies, particularly those based on diffusion models, have shown promise in addressing these limitations. In this work, we introduce a novel latent diffusion based pretraining framework, CrysLDNet, designed to mitigate data scarcity. Our approach integrates a Variational Autoencoder (VAE) with a diffusion model during the pretraining stage. The VAE encoder maps 3D crystal structures into a smooth latent space within which the diffusion process is applied. This latent diffusion pretraining enables the graph encoder to effectively capture structural and chemical semantics from large-scale unlabeled data, which can then be finetuned for specific property prediction tasks. Comprehensive experiments on popular DFT datasets for property prediction reveal that CrysLDNet significantly outperforms both training-from-scratch and pretrained baselines, with improvements of 4.26% and 4.90% on the JARVIS and MP datasets, respectively. Additionally, the learned representations remain robust in sparse-data conditions and are expressive enough to correct DFT errors when finetuned with limited experimental data. Code is available at: https://github.com/shrimonmuke0202/CrysLDNet.git. --- ### GIRL-DETR: Gradient-Isolated Reinforcement Learning for Video Moment Retrieval Authors: Shihang Zhang, Mingjin Kuai, Ye Wei, Zhen Zhang, Wei Ji Published: 2026-05-30 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.00775v1 Video Moment Retrieval (VMR) task requires accurately localizing temporal boundaries aligned with natural language queries, but many models suffer from a misalignment between continuous surrogate losses and non-differentiable metrics, leading to optimization stagnation during the late stages of training and trapping boundary predictions in suboptimal solutions. Although Reinforcement Learning (RL) post-training successfully optimizes localization results for large models, applying it directly to lightweight networks easily disrupts the fragile feature representations established during the supervised phase. To overcome this optimization bottleneck, we propose Gradient-Isolated Reinforcement Learning for DETR (GIRL-DETR), introducing RL post-training into a lightweight temporal localization framework for the first time. The input video and text features first establish early alignment through Cross-Modal Interaction (CMI) before entering the transformer encoder. Subsequently, a Text-Guided Gating (TGG) mechanism dynamically injects semantic priors into the queries before the transformer decoder generates candidate proposals, providing high signal-to-noise ratio inputs for temporal prediction. After the supervised training reaches convergence, the backbone network is frozen to protect the feature manifold, while the detection head directly optimizes the non-differentiable evaluation metric tIoU to enhance localization accuracy through a Three-stage Progressive Reinforcement Learning (TPRL) strategy. This approach achieves an orthogonal decoupling of state representation and metric optimization. Experiments on Charades-STA, QVHighlights, and TACoS demonstrate that GIRL-DETR effectively resolves surrogate loss degradation and achieves substantial accuracy improvements with minimal parameter updates, providing a robust new pathway for RL applications in lightweight VMR models. --- ### Logit Distillation on Manifolds: Mapping by Learning Authors: Yiru Yang, Junling Wang, Nishant Kumar Singh, Luohong Wu, Haoran Yan Published: 2026-05-30 Categories: cs.LG, cs.AI, cs.SD arXiv: https://arxiv.org/abs/2606.00771v1 A simple way to improve the performance of almost any machine learning model is not to train a single but several models with diverse algorithms which will make slightly distinct kinds of predictions and errors on the same data, and thus improve the average predictions and robustness. However, making predictions using a whole ensemble of models is cumbersome and computationally too expensive to allow deployment to a large number of users, especially if the models are large neural nets. In response to this, we introduce a layer and point wise projection mapping, which maps student and teacher representations into an aligned high-dimensional embedding space during training process. The proposed approach combined with LoRA injection reduces the student model trainable parameters to less than 1% of the teacher model, while significantly improving word error rate (WER) compared to other distillation methods, as demonstrated in ablation studies. Unlike a mixture of experts, our method can be trained rapidly and in parallel. --- ### FALAT: Tracing Failures in LLM Agent Trajectories via Dependency-Guided Search Authors: Md Nakhla Rafi, Md Ahasanuzzaman, Dong Jae Kim, Zhijie Wang, Tse-Hsun Chen Published: 2026-05-30 Categories: cs.AI arXiv: https://arxiv.org/abs/2606.00765v1 LLM-based agents increasingly solve complex tasks through long trajectories involving reasoning steps, tool calls, and inter-agent communication. However, when these agents fail, it is often unclear which agent caused the failure and which step introduced the decisive error. This attribution problem is challenging because mistakes can propagate across the trajectory: later actions may appear incorrect, but only because they depend on an earlier corrupted state. Therefore, failure attribution cannot be treated as independent step-level classification. We propose FALAT, a diagnostic framework for failure attribution in LLM agent trajectories. FALAT frames attribution as a dependency-guided search problem. It first constructs an expectation of how the task should be solved and uses this expectation to identify suspicious regions in the trajectory. It then traces dependencies among decisions, tool outputs, and agent messages to distinguish error-introducing steps from steps that merely inherit or propagate prior mistakes. Finally, FALAT evaluates whether correcting a candidate step would be sufficient to recover the expected outcome, allowing it to identify both the responsible agent and the decisive failure step. We evaluate FALAT on the Who&When benchmark, which includes both algorithm-generated and hand-crafted multi-agent failure trajectories. The results show that FALAT consistently improves responsible-agent and decisive-step attribution. Its best configurations achieve 46.0% step-level accuracy on algorithm-generated trajectories and 29.1% on the more challenging hand-crafted trajectories, outperforming specialized attribution baselines and direct prompting with standalone LLMs. These findings suggest that dependency-aware reasoning is essential for reliable failure diagnosis in LLM agent systems. --- ### Confidence-Adaptive SwiGLU for Mixture-of-Experts Authors: Shaohua Li, Xiuchao Sui, Xiaobing Sun, Yuhang Wu, Liangli Zhen, Yong Liu, Rick Siow Mong Goh Published: 2026-05-30 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2606.00761v1 SwiGLU has become a standard gated activation in modern Transformer MLPs, yet its gate sharpness -- the smoothness and selectivity of the gating function -- is typically fixed throughout training. In this work, we propose Confidence-Aware SwiGLU ($κ$-SwiGLU), a variant of SwiGLU for Mixture-of-Experts (MoE) models that adjusts expert gate sharpness according to token-level routing confidence. Specifically, $κ$-SwiGLU parameterizes the SiLU gate sharpness coefficient as a learnable function of the router logit, enabling each expert gate unit to interpolate between smooth, broadly active gating and sharp, selective gating. We evaluate $κ$-SwiGLU on the FineWeb-Edu dataset across MoE Transformer models ranging from 8 to 28 layers. Across these settings, $κ$-SwiGLU improves mean CORE performance while adding negligible parameters and incurring only a small computational overhead, demonstrating that confidence-aware gate sharpness is a promising mechanism for improving MoE MLPs. The code is available at https://github.com/askerlee/kappa-swiglu. --- ### Distributed GNEP Algorithms without Multiplier Sharing and Applications to Multi-Robot Coordination and Contextual Bandit-Based Active Learning Authors: Shao-An Yin Published: 2026-05-30 Categories: cs.LG arXiv: https://arxiv.org/abs/2606.00759v1 Recent advances in artificial intelligence have expanded the focus from classical optimization to include equilibrium analysis in noncooperative games. Many such games involve shared constraints, leading to Generalized Nash Equilibrium Problems (GNEPs). Existing distributed algorithms typically require agents to exchange Lagrange multipliers to enforce consensus and compute variational-GNEs (v-GNEs). This work introduces fully distributed continuous-time algorithms and establishes convergence without requiring multiplier exchange, thereby reducing information exchange per iteration while improving privacy preservation. The analysis focuses on strongly monotone games with convex individual constraints and linear shared constraints. I also propose several discretization schemes for the continuous-time algorithms. The proposed approach converges to general GNEs, rather than being restricted to v-GNEs, with the attained equilibrium depending on the initialization. The effectiveness of the proposed method is demonstrated through applications in multi-robot coordination and placement. In the second part, this work includes research conducted in collaboration with Amazon scientists. One of the most challenging problems in real-world machine learning is labeled data collection, which typically requires substantial human effort and cost. Active learning aims to reduce this labeling requirement. Existing handcrafted active learning strategies, however, generally perform well only on specific types of datasets, which are often unknown in advance. In this work, I propose using contextual bandits to adaptively select the most suitable active learning strategy. The effectiveness of the proposed approach is demonstrated on publicly available external datasets. --- ### Statistical Testing on Directed Graphs by Surrogate Data Generation Authors: Chun Hei Michael Chan, Alexandre Cionca, Dimitri Van De Ville Published: 2026-05-30 Categories: stat.ML, cs.LG, eess.SP, stat.ME arXiv: https://arxiv.org/abs/2606.00758v1 In recent years, graph signal processing has emerged as a powerful framework at the intersection of signal processing and graph theory, providing tools for the analysis of signals defined on nodes while accounting for their relationships represented by edges. These tools have been successfully applied to various settings, including statistical hypothesis testing. In particular, non-parametric approaches based on surrogate generation have been proposed for signals on undirected graphs. However, they are yet to be extended to directed graphs. In this work, we first revisit the notion of stationary graph signals on directed graphs. Specifically, and through the eigendecomposition of the graph shift operator, we define directed graph wide-sense stationary signals. Then, we propose a new framework to generate surrogate graph signals that preserve covariance structure under stationarity assumptions. Null distributions of the test metric can then be constructed from these surrogates and serve as a reference for the empirical data. Finally, we provide guiding examples and an application on real data, in which we compare the performance of our framework with existing techniques for undirected graphs or based on naive permutation, demonstrating feasibility and superiority of the proposed approach. --- ### RADE: Random Add-Drop Edge as a Regularizer Authors: Danial Saber, Amirali Salehi-Abari Published: 2026-05-30 Categories: cs.LG arXiv: https://arxiv.org/abs/2606.00757v2 Graph Neural Networks (GNNs) suffer from overfitting and over-squashing of long-range information. Stochastic graph augmentations (e.g., edge deletion) regularize training against overfitting but can introduce train-inference misalignment and do not improve over-squashing. In contrast, rewiring methods improve connectivity to mitigate over-squashing, but are not designed to regularize training. We propose Random Add-Drop Edge (RADE), a stochastic graph augmentation method that jointly drops and adds edges to address both overfitting and over-squashing simultaneously. RADE is provably designed to align training and inference so that random augmentations regularize training without distribution shift, while supporting long-range communication at inference. We further propose and study a mini-batch gradient-norm balancing algorithm that adapts deletion and addition rates during training, rendering RADE hyperparameter-free in practice. Experiments on node- and graph-classification benchmarks show that RADE is a strong regularizer and mitigates over-squashing. Ablations support the roles of train-inference alignment, adaptive rate selection, and the complementary effects of random edge deletion and edge addition. --- ### CoMIC: Collaborative Memory and Insights Circulation for Long-Horizon LLM Agents in Cloud-Edge Systems Authors: Yannan Wang, Longli Yang, Zhen Liu, Abhishek Kumar, Carsten Maple Published: 2026-05-30 Categories: cs.AI arXiv: https://arxiv.org/abs/2606.00756v1 Deploying lightweight Large Language Model (LLM) agents on edge servers can reduce latency and move agentic services closer to users, but resource-constrained edge models often struggle with long-horizon tasks that require persistent memory, subgoal tracking, and reflection. Fine-tuning edge models after deployment is costly and difficult to scale across heterogeneous nodes, while purely local memory leaves agents with isolated experience and growing prompt context. We propose \textsc{CoMIC}, a parameter-update-free cloud-edge framework for Collaborative Memory and Insights Circulation. \textsc{CoMIC} follows a \textit{Centralized Reflection, Decentralized Execution} design: edge agents execute locally using subgoal-oriented hierarchical memory and selective re-expansion of relevant histories, while a cloud-side LLM critic asynchronously evaluates completed trajectories, filters reusable experience, and aggregates cross-agent guidance keyed by semantic subgoal identifiers. Across five long-horizon agent tasks spanning symbolic planning and text interaction, \textsc{CoMIC} improves progress rate and action grounding for weak edge agents and yields task-dependent success-rate gains without updating model parameters. --- ### Internalize the Temperature: On-Policy Self-Distillation as Policy Reheater for Reinforcement Learning Authors: Xuewei Yang, Jiachen Yu, Jie Wu, Shaoning Sun, Junjie Wang, Yujiu Yang Published: 2026-05-30 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2606.00755v2 Reinforcement learning from verifiable rewards improves the reasoning ability of large language models, but often suffers from entropy collapse, in which increasingly concentrated policies reduce rollout diversity and useful learning signals. Existing remedies either constrain the RL objective (e.g., entropy regularization) or adjust sampling temperature during rollout collection, but these interventions remain external to the model parameters. We propose Temperature-Scaled On-Policy Self-Distillation (TS-OPSD), a lightweight policy reheating method that internalizes the exploratory effect of temperature into model parameters. Starting from an entropy-collapsed RL checkpoint, TS-OPSD constructs a self-teacher by applying high-temperature scaling to the model's own logits, then distills the resulting smoother distribution back into the student. This policy reheating requires no external teacher, privileged data, or additional inference cost. Experiments on Qwen3-4B-Base and Qwen3-8B-Base show that policy reheating yields a stronger initialization for continued RL than both standard continued RL and rollout-level temperature reheating. Further analyses show that TS-OPSD mainly reduces output sharpness while preserving intermediate representations, top candidate sets, and reasoning capability. These results suggest that entropy restoration can serve as a simple post-collapse intervention for extending reasoning-oriented RL. --- ### Internalize the Temperature: On-Policy Self-Distillation as Policy Reheater for Reinforcement Learning Authors: Xuewei Yang, Jiachen Yu, Jie Wu, Shaoning Sun, Junjie Wang, Yujiu Yang Published: 2026-05-30 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2606.00755v1 Reinforcement learning from verifiable rewards improves the reasoning ability of large language models, but often suffers from entropy collapse, in which increasingly concentrated policies reduce rollout diversity and useful learning signals. Existing remedies either constrain the RL objective (e.g., entropy regularization) or adjust sampling temperature during rollout collection, but these interventions remain external to the model parameters. We propose Temperature-Scaled On-Policy Self-Distillation (TS-OPSD), a lightweight policy reheating method that internalizes the exploratory effect of temperature into model parameters. Starting from an entropy-collapsed RL checkpoint, TS-OPSD constructs a self-teacher by applying high-temperature scaling to the model's own logits, then distills the resulting smoother distribution back into the student. This policy reheating requires no external teacher, privileged data, or additional inference cost. Experiments on Qwen3-4B-Base and Qwen3-8B-Base show that policy reheating yields a stronger initialization for continued RL than both standard continued RL and rollout-level temperature reheating. Further analyses show that TS-OPSD mainly reduces output sharpness while preserving intermediate representations, top candidate sets, and reasoning capability. These results suggest that entropy restoration can serve as a simple post-collapse intervention for extending reasoning-oriented RL. --- ### Causal Density Functions Authors: Sridhar Mahadevan Published: 2026-05-30 Categories: stat.ME, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.00754v2 We study the full density ratio between a specified intervention regime $P_a$ and an observational regime $P_0$, $ρ_a=dP_a/dP_0$, under the prerequisite $P_a\ll P_0$. We call the regime-indexed ratio a causal density function when $P_a$ is an identified or directly observed interventional law. The underlying Radon-Nikodym derivative and the identity $ \mathbb{E}_{a}(f(Z)) = \mathbb{E}_{0}\!\left(f(Z)ρ_a(Z)\right) $ are classical importance weighting, not new identification results. Our narrower question is whether retaining the entire pointwise ratio is useful as a reusable diagnostic across several downstream functionals. We evaluate a two-density plug-in baseline through held-out moment transport and overlap stress tests on synthetic and perturbation data. We also report a pairwise graph-scoring heuristic as a negative result: its F1 is \(0.10\) on a synthetic DAG, \(0.12\) on Sachs, and \(0.33\) on a multi-regime chain. These experiments do not establish an estimation advantage over direct density-ratio, inverse-probability, Riesz, or doubly robust methods; they instead delimit what the pointwise ratio target and the present plug-in estimator do and do not provide. --- ### A multimodal dataset of photoplethysmography and continuous behavioral responses to ASMR and nature videos Authors: Tushar Das, Daigo Hozaki, Koushlendra Kumar Singh, Hirohito M. Kondo Published: 2026-05-30 Categories: cs.LG, cs.CE, cs.HC arXiv: https://arxiv.org/abs/2606.00752v1 Autonomous Sensory Meridian Response (ASMR) is a somatosensory phenomenon characterized by pleasant tingling sensations and cardiovascular slowing. However, ASMR research has been hindered by a dearth of standardized, open-access multimodal datasets. To address this limitation, we present REST-ASMR (Response to Environmental & Sensory Triggers), a synchronized multimodal dataset designed to capture behavioral reports and physiological dynamics during ASMR, with nature-relaxation videos as control stimuli. The dataset includes high-resolution photoplethysmography (PPG), time-aligned audiovisual stimuli, and continuous subjective annotations from 34 participants. Technical validation showed high stimulus efficacy (97% responder rate), significant stimulus-specific inter-subject agreement (p < 0.05), and a robust PPG-derived ASMR-specific cardiovascular deceleration. Additionally, a Bidirectional Long-Short Term Memory model successfully predicted subjective ASMR tingle states, achieving video-level ASMR vs. Nature classification with perfect accuracy and a frame-level global mean accuracy of 75.51%, macro F1-score of 71.86%, and 100% Nature-baseline specificity, under a strict, leakage-free subject-video double-independent 4-fold cross-validation. REST-ASMR constitutes a dense temporal foundation for affective computing, multimodal research, and the development of personalized models of relaxation-related responses. --- ### Head-Pose-Aware Visual Speech Recognition with FiLM Modulation Authors: Matthew Kit Khinn Teng, Haibo Zhang, Takeshi Saitoh Published: 2026-05-30 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.00751v2 Visual Speech Recognition (VSR) aims to recognize speech from visual cues such as lip movements. Still, its performance is fundamentally limited by viseme ambiguity and pose-induced variations that introduce geometric distortions and occlusions. Existing approaches mainly rely on linguistic context or implicit invariance, leaving visual representations insufficiently robust under non-frontal views. In this work, we propose a pose-aware phoneme-level framework, termed HP-VSR-ResFiLM, that explicitly incorporates head-pose information into visual feature extraction. The proposed framework adopts a two-stage pipeline consisting of a pose-conditioned visual encoder in Stage~1 and a pretrained NLLB language model in Stage~2 for phoneme-to-text reconstruction. Specifically, Stage~1 incorporates a pose-conditioned residual Feature-wise Linear Modulation (FiLM) block after the 2D CNN frontend to refine visual representations adaptively using head-pose information. Experiments on LRS2 and LRS3 demonstrate that HP-VSR-ResFiLM achieves competitive performance under comparable training conditions, attaining word error rates (WER) of 24.7% and 30.3%, respectively, without relying on additional training data. Comprehensive ablation studies reveal that different FiLM modulation strategies exhibit complementary strengths: HP-VSR-FiLMFuse (L4) achieves the best overall performance on LRS2, whereas the proposed HP-VSR-ResFiLM provides the greatest improvements under large head-pose variations on LRS3 and consistently outperforms previous pose-aware methods for high-yaw samples. These results demonstrate the effectiveness of explicit pose-conditioned feature modulation for robust visual speech recognition in unconstrained settings. --- ### Head-Pose-Aware Visual Speech Recognition with FiLM Modulation Authors: Matthew Kit Khinn Teng, Haibo Zhang, Takeshi Saitoh Published: 2026-05-30 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.00751v1 Visual Speech Recognition (VSR) aims to recognize speech from visual cues such as lip movements, but its performance is fundamentally limited by viseme ambiguity and pose-induced variations that introduce geometric distortions and occlusions. Existing approaches mainly rely on linguistic context or implicit invariance, leaving visual representations insufficiently robust under non-frontal views. In this work, we propose a pose-aware phoneme-level framework, termed HP-VSR-ResFiLM, that explicitly incorporates head-pose information into visual feature extraction. The proposed framework adopts a two-stage pipeline consisting of a pose-conditioned visual encoder in Stage 1 and a pretrained NLLB language model in Stage 2 for phoneme-to-text reconstruction. Specifically, Stage 1 incorporates a pose-conditioned residual Feature-wise Linear Modulation (FiLM) block after the 2D CNN frontend to adaptively refine visual representations using head-pose information. Experiments on LRS2 and LRS3 demonstrate that HP-VSR-ResFiLM achieves competitive performance under comparable training conditions, attaining word error rates (WER) of 25.0% and 33.2%, respectively, without relying on additional training data. Ablation studies further show that a single residual FiLM block consistently improves overall WER, while deeper modulation at Layers 3 and 4 provides larger gains for samples with yaw angles greater than 30° without degrading performance for smaller pose variations. These findings demonstrate that explicit pose-aware feature modulation offers an effective and computationally efficient solution for improving VSR robustness in unconstrained settings. --- ### I-WebGenBench : Evaluating Interactivity in LLM-Generated Scientific Web Applications Authors: Dasen Dai, Biao Wu, Meng Fang, Shuoqi Li, Wenhao Wang Published: 2026-05-30 Categories: cs.CL arXiv: https://arxiv.org/abs/2606.00750v1 Recent advances in visual language models have enabled autonomous agents for complex reasoning, tool use, and document understanding. However, existing document agents mainly transform papers into static artifacts such as summaries, webpages, or slides, which are insufficient for technical papers involving dynamic mechanisms and state transitions. In this work, we propose a Paper-to-Interactive-System Agent that converts research papers into executable interactive web systems. Given a PDF paper, the agent performs end-to-end processing without human intervention, including paper understanding, system modeling, and interactive webpage synthesis, enabling users to manipulate inputs and observe dynamic behaviors. To evaluate this task, we introduce a benchmark of 19 research papers paired with expert-built interactive systems as ground truth. We further propose PaperVoyager, a structured generation framework that explicitly models mechanisms and interaction logic during synthesis. Experiments show that PaperVoyager significantly improves the quality of generated interactive systems, offering a new paradigm for interactive scientific paper understanding. --- ### SkyShield: Occupancy as a Safety Interface for Low-Altitude UAV Autonomy Authors: Jie Gao, Jie Ma, Kaihui Lin, Kai Ye, Miaohui Zhang, Pingyang Dai, Liujuan Cao Published: 2026-05-30 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.00747v2 For low-altitude Unmanned Aerial Vehicle (UAV) autonomy, 3D spatial understanding is not merely a perception objective, but the safety interface between human instructions and physical flight. In human-scale urban airspace below 20 meters, thin geometry, occlusions, vegetation, and urban clutter define whether an aerial agent can safely enter the space ahead. However, existing UAV datasets mainly provide 2D annotations or 3D boxes, while driving-oriented occupancy benchmarks assume stable ground-level sensor rigs. Both miss the defining regime of low-altitude flight: a front-facing monocular camera observing occupied and free space from a moving aerial body with frame-wise changing 6-DoF pose and camera extrinsics. To bridge this gap, we introduce SkyShield, to the best of our knowledge the first front-view monocular semantic occupancy benchmark for urban UAV flight below 20 meters. Built on CARLA, SkyShield contains 36K front-view UAV samples across diverse urban scenes and weather conditions, pairing each image with frame-wise 6-DoF UAV pose, frame-wise dynamic camera geometry, UAV states, and front-frustum semantic occupancy labels. We further propose KAR-mIoU, a UAV-centric and dynamics-aware metric that re-weights voxel-level evaluation by kinematic reachability and time-to-collision, revealing safety-critical risks hidden by conventional mIoU. To tackle this challenging new setting, we provide SkyOcc, a geometry-first monocular baseline that integrates frame-wise UAV attitude into projection, fuses temporal occupancy features, and applies safety-prior optimization to preserve sparse collision-critical structures. Together, SkyShield, KAR-mIoU, and SkyOcc establish occupancy as a safety interface for low-altitude aerial autonomy. Code and dataset will be released publicly. --- ### Scaling Parallel Sequence Models to Foundation-Scale Vision Encoders Authors: Yitong Jiang, Hongjun Wang, Collin McCarthy, Hanrong Ye, David Wehr, Xinhao Li, Qi Dou, Tianfan Xue, Ka Chun Cheung, Simon See, Wonmin Byeon, Ke Chen et al. Published: 2026-05-30 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2606.00746v1 Vision foundation models are bottlenecked by the quadratic cost of self-attention, which limits usable resolution and increases the cost of large-scale pretraining. Subquadratic alternatives such as linear attention and state-space models reduce this cost, but often serialize images into 1D token streams and weaken the 2D spatial structure important for vision. Generalized Spatial Propagation Networks (GSPN) instead propagate context directly on the 2D grid through line-scan recurrences, achieving near-linear complexity without positional embeddings, but have seen little use as foundation-scale encoders. We present C-GSPN, a foundation-scale vision encoder based on 2D spatial propagation. C-GSPN makes the operator practical through three improvements: (1) a fast GSPN CUDA kernel that fuses per-step launches into a single warp-specialized implementation with shared-memory tiling, coalesced access, and a compact multi-channel propagation, reaching over 90% of peak memory bandwidth and running up to 40--52x faster than the original GSPN implementation; (2) a compressed latent-space propagation block with fused normalization, which turns kernel-level speed into block- and model-level efficiency; and (3) a two-stage cross-operator distillation recipe that trains the new architecture from an attention teacher without the cost of from-scratch foundation-scale training. Distilled with 600M image-text pairs, C-GSPN matches an isomorphic ViT baseline with 15% fewer parameters, improves ADE20K segmentation by +2.1%, transfers to high resolution with a fraction of the data needed from scratch, and delivers a 4x end-to-end block speedup at 2K with single-pass, tiling-free inference. --- ### Quantum Tunneling-Aware Machine Learning: Physics-Derived Noise Models for Robust Deployment Authors: Uiwon Hwang, Jaeho Hwang Published: 2026-05-30 Categories: cs.LG, cs.AI, stat.ML arXiv: https://arxiv.org/abs/2606.00741v1 Transistor scaling is approaching a quantum-mechanical limit, as thin gate oxides induce electron leakage through quantum tunneling. Unlike conventional digital systems, AI inference can tolerate such errors provided their structure is modeled correctly. In this paper, we introduce quantum tunneling-aware machine learning (QTAML). We derive the deployment-time weight-error distribution from first principles using the Wentzel-Kramers-Brillouin (WKB) approximation and show that it has structure that generic Gaussian noise models miss: an exact affine mean drift, a per-bit variance hierarchy dominated by the most-significant bit, and a per-layer dependence on $\|W_\ell\|_\infty$ and the trained-network Jacobian. We package these three structural properties into a single deployment-time algorithm, Tunneling-Aware Compensation (TAC), that combines closed-form mean correction with an optimal layer-adaptive bit-budget allocation derived from the WKB variance decomposition. Across four convolutional architectures at $p_\mathrm{flip}$=0.10 and a transformer encoder at $p_\mathrm{flip}$=0.05, TAC reaches $95\%$ of clean accuracy with 3.4$\times$ to 33.6$\times$ less ECC overhead than Uniform-MSP, the natural baseline derived from the same physics. The closed-form saturation ratio $ρ^*$ predicts these gains in advance, and on heterogeneous architectures WKB-derived scoring outperforms magnitude-based allocation by up to 24 percentage points at small budgets. The algorithm requires no retraining, no labels, and no inference-time overhead. We also verify the WKB-derived distributional theorems to Monte Carlo precision. These results connect WKB tunneling physics with noise-aware deep learning and suggest a principled path toward hardware--software co-design beyond conventional scaling limits. --- ### Score $\times$ Decoder: A Unified View of Unsupervised Inference-Time Scaling for Hallucination Mitigation Authors: Yun-Chen Cheng, Che-Yu Lin, Cheng-Lin Yang Published: 2026-05-30 Categories: cs.LG arXiv: https://arxiv.org/abs/2606.00739v1 Large language models hallucinate even when the answer lies within their parameters. While inference-time scaling can surface this latent knowledge, the most effective methods require supervision: a trained verifier or reward model. We ask what can be done with only a base language model: which intrinsic signal best identifies correct outputs, and how should it be decoded? We cast this as a score~$\times$~decoder grid pairing four scores (perplexity, contrastive, power-distribution likelihood, and self-verification) with three decoding families (optimization, sampling, consensus), and evaluate every cell on MATH500 with the base and instruction-tuned Qwen3-1.7B. While self-verification, which prompts the model to judge its own answer and is sharpened by a training-free virtual-thinking prefix, works well in most settings, no score has a fixed quality: its value depends on the decoder that consumes it and on model capability. When no supervision is available, the score and the decoding family must be chosen together. --- ### SORA: Free Second-Order Attacks in Fast Adversarial Training Authors: Mazdak Teymourian, Ramtin Moslemi, Farzan Rahmani, Mohammad Hossein Rohban Published: 2026-05-30 Categories: cs.LG, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2606.00738v1 Adversarial Training (AT) is a leading defense against adversarial examples but often suffers from Catastrophic Overfitting (CO) in efficient single-step variants, where robustness to multi-step attacks collapses despite high single-step performance. We address this failure mode with two contributions. First, we formalize Epsilon Overfitting (EO), a perspective in which fixed perturbation magnitudes and directions exacerbate CO, and show that introducing perturbation variability significantly improves robust generalization across different architectures and datasets. Second, we propose PertAlign (Perturbation Alignment), a theoretically grounded, computationally negligible metric that predicts CO onset by measuring gradient alignment across attack stages. Leveraging these insights, we introduce SORA, an adaptive step-size AT method that dynamically adjusts perturbations based on loss surface geometry. SORA consistently prevents CO, achieves state-of-the-art robustness and clean accuracy, and generalizes across datasets and architectures using a single fixed set of hyperparameters, which is essential for applicability in fast AT. Extensive experiments on diverse datasets and architectures show that SORA matches or surpasses the robustness of prior methods while delivering higher clean accuracy and superior efficiency. Code is available at https://github.com/SecondOrderAT/SORA. --- ### ViBE: Co-Optimizing Workload Skew and Hardware Variability for MoE Serving Authors: Seokjin Go, Marko Scrbak, Ephrem Wu, Srilatha Manne, Divya Mahajan Published: 2026-05-30 Categories: cs.DC, cs.LG arXiv: https://arxiv.org/abs/2606.00735v1 In distributed Mixture-of-Experts (MoE) inference, input-dependent token routing interacts with GPU performance variability to create persistent stragglers under synchronized execution, where the slowest GPU determines layer latency. This performance variability is inherent to modern accelerators: manufacturing variation, power limits, and thermal conditions introduce measurable execution-time differences across nominally identical GPUs. The core challenge is that MoE execution-time imbalance arises from the interaction of workload skew and hardware asymmetry. Token routing produces uneven and layer-varying expert loads, while GPU throughput depends on device-specific operating characteristics and workload intensity. Prior work mitigates routing skew but assumes homogeneous hardware, optimizing token balance rather than execution latency. As a result, even balanced token assignments can leave hardware-induced stragglers unaddressed. Thus, we propose Variability-Informed Binning of Experts (ViBE), a hardware-aware expert placement framework that minimizes execution-time imbalance across GPUs. ViBE combines per-GPU performance modeling with expert activation profiling to assign high-load experts to faster devices and low-load experts to slower ones, reducing layer-level stragglers without modifying model semantics or hardware. Because both workload characteristics and effective GPU throughput can shift across serving conditions, ViBE supports lightweight recalibration under workload/performance drift to refresh its routing and performance estimates when needed. Results show that ViBE consistently reduces execution-time imbalance and improves SLO attainment by 14%, while lowering P90 TTFT by up to 45%. We further show that the impact of hardware variability increases at scale, making variability-aware placement important for efficient, high-utilization LLM serving. --- ### SHARP: Sleep-based Hierarchical Accelerated Replay for Long Range Non-Stationary Temporal Pattern Recognition Authors: Jayanta Dey, Shikhar Srivastava, Itamar Lerner, Christopher Kanan, Dhireesha Kudithipudi Published: 2026-05-30 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.00732v4 Learning long-range non-stationary temporal patterns remains a core challenge for modern sequence models, particularly in strict streaming settings. In these settings, data arrive sequentially and must be processed in a single pass without simultaneously revisiting past observations. Standard architectures, including recurrent neural networks and transformers, are constrained by either truncated backpropagation through time horizon or explicit input window length for long range credit assignment. To address these limitations, we propose SHARP (Sleep-based Hierarchical Accelerated Replay), a framework that decomposes temporal learning into two complementary components: a memory module that accumulates a structured history of past inputs, and a pattern-recognition module that operates over this memory. This separation enables resource- and compute-efficient adaptation to non-stationary dynamics by eliminating the need for backpropagation through time across many steps for long-range credit assignment. Inspired by the accelerated replay observed in rodents during slow-wave sleep, SHARP incorporates offline (sleep) phases in which temporally structured memory traces are replayed in an accelerated form and integrated into higher-level memory representations, improving long-range context retention. Through controlled simulations and ablation studies, we characterize the key properties of the proposed framework. In benchmark datasets such as text8 and PG-19, we demonstrate that SHARP improves over recurrent baselines by retaining next-token predictive performance on previously seen data while continuing to learn from the current stream and generalizing to future unseen data. These gains are enabled by its hierarchical structure, which yields an exponentially increasing effective temporal context with only linear-time computational cost. --- ### AI Sovereignty as National Learning Capacity: A Human-Centered Learning Mechanics Viewpoint on France, the United States, and China Authors: Kim Phuc Tran Published: 2026-05-30 Categories: cs.AI arXiv: https://arxiv.org/abs/2606.00729v2 Artificial intelligence in France is often discussed through separate dimensions such as investment, compute, regulation, employment, sovereignty, and education. This viewpoint paper proposes a unified interpretation: France can be analyzed as a national AI learning system. Building on Human-Centered Learning Mechanics (HCLM), we use HCLM not as a validated econometric model, but as a conceptual and diagnostic lens for interpreting national AI development as a balance between information injection, absorptive capacity, and institutional dissipation. Information injection includes compute, data, talent, research, capital, industrial deployment, and policy experimentation. Institutional dissipation refers to avoidable frictions such as administrative overload, coordination failures, energy constraints, regulatory uncertainty, talent mobility pressures, and weak industrial absorption. Regulation is not treated as mere friction: adaptive governance, trusted data spaces, and safety-oriented standards may increase long-term learning capacity by improving legitimacy, interoperability, and social trust. The central claim is not that a country follows neural-network equations, but that AI sovereignty depends on how effectively it converts distributed information into absorbed, coordinated, and socially legitimate capability. The paper connects HCLM with neural scaling laws, endogenous growth theory, creative destruction, absorptive capacity, and coordination mechanisms. It offers a formal heuristic, policy indicators, illustrative scenarios, and implications for France. The numerical results are diagnostic scenarios, not econometric estimates or official rankings. The proposed viewpoint reframes AI policy as the governance of an open, strategic, non-equilibrium learning system that should be tested with historical and cross-country data. --- ### From Empathy to Personalized Empathy: Adapting Empathetic Strategies to Individual Users Authors: Wuqiang Zheng, Chengbing Wang, Yilin Yang, Junyi Cheng, Jianfei Xiao, Hu Sun, Yi Xie, Yangyang Li, Wenjie Wang Published: 2026-05-30 Categories: cs.CL arXiv: https://arxiv.org/abs/2606.00728v2 As Large Language Models (LLMs) are increasingly deployed in long-term interactions with users, empathy has become an increasingly important capability. However, existing research overlooks the influence of users' personality traits on empathetic strategies during long-term interactions. To address this gap, we introduce the task of personalized empathy, which focuses on adapting empathetic strategies according to users' personalized characteristics derived from history. To study and enhance this capability, we construct PersonaEmp, a personalized empathy dataset built from long-term user-AI interactions, featuring rich user histories, persona information, and empathy-seeking queries. We further propose PereGRM, a reward modeling framework that combines the empathy evaluation structure with dynamic evaluation criteria generation for fine-grained reward modeling. Experimental results across different settings and multiple judge models show that PereGRM consistently achieves the strongest performance improvements, indicating its effectiveness for enhancing personalized empathetic capabilities. --- ### From Empathy to Personalized Empathy: Adapting Empathetic Strategies to Individual Users Authors: Wuqiang Zheng, Chengbing Wang, Yilin Yang, Junyi Cheng, Jianfei Xiao, Hu Sun, Yi Xie, Yangyang Li, Wenjie Wang Published: 2026-05-30 Categories: cs.CL arXiv: https://arxiv.org/abs/2606.00728v1 As Large Language Models (LLMs) are increasingly deployed in long-term interactions with users, empathy has become an increasingly important capability. However, existing research overlooks the influence of users' personality traits on empathetic strategies during long-term interactions. To address this gap, we introduce the task of personalized empathy, which focuses on adapting empathetic strategies according to users' personalized characteristics derived from history. To study and enhance this capability, we construct PersonaEmp, a personalized empathy dataset built from long-term user-AI interactions, featuring rich user histories, persona information, and empathy-seeking queries. We further propose PereGRM, a reward modeling framework that combines the empathy evaluation structure with dynamic evaluation criteria generation for fine-grained reward modeling. Experimental results across different settings and multiple judge models show that PereGRM consistently achieves the strongest performance improvements, indicating its effectiveness for enhancing personalized empathetic capabilities. --- ### Latent Reward Steering: An Adaptive Inference-Time Framework that Implicitly Promotes Cognitive Behaviors in Reasoning LLMs Authors: Jiakang Li, Guanyu Zhu, Can Jin, Chenxi Huang, Dexu Yu, Ronghao Chen, Yang Zhou, Hongwu Peng, Xuanqi Lan, Dimitris N. Metaxas, Youhua Li Published: 2026-05-30 Categories: cs.AI arXiv: https://arxiv.org/abs/2606.00726v3 Strong reasoning depends not only on model knowledge but also on how effectively cognitive behaviors are deployed during generation. Existing methods often rely on explicit behavior-level control, making them insufficiently adaptive when failures and required corrections vary across reasoning states, tasks, and models. To this end, we propose Latent Reward Steering (LRS), an adaptive inference-time framework that promotes cognitive behaviors by optimizing the sparse-autoencoder (SAE) latent states that implicitly carry them. Rather than relying on predefined cognitive behaviors or steering directions derived from them, LRS trains a latent reward model on reasoning traces by final answer correctness to estimate the quality of intermediate latent states. During inference, reward gradients provide state-specific correction directions for fragile latent states, while a reward and confidence gate restricts intervention to states the reward signal flags as fragile. Experiments on multiple reasoning LLM backbones and benchmarks show that \ours consistently improves performance over various baselines, and post-hoc analyses further indicate that \ours implicitly promotes good cognitive behaviors that fix the original reasoning errors. Code is available at: https://github.com/jiakanglee/Latent-Reward-Steering. --- ### Latent Reward Steering: An Adaptive Inference-Time Framework that Implicitly Promotes Cognitive Behaviors in Reasoning LLMs Authors: Jiakang Li, Guanyu Zhu, Can Jin, Chenxi Huang, Dexu Yu, Ronghao Chen, Yang Zhou, Hongwu Peng, Xuanqi Lan, Dimitris N. Metaxas, Youhua Li Published: 2026-05-30 Categories: cs.AI arXiv: https://arxiv.org/abs/2606.00726v2 Strong reasoning depends not only on model knowledge but also on how effectively cognitive behaviors are deployed during generation. Existing methods often rely on explicit behavior-level control, making them insufficiently adaptive when failures and required corrections vary across reasoning states, tasks, and models. To this end, we propose Latent Reward Steering (LRS), an adaptive inference-time framework that promotes cognitive behaviors by optimizing the sparse-autoencoder (SAE) latent states that implicitly carry them. Rather than relying on predefined cognitive behaviors or steering directions derived from them, LRS trains a latent reward model on reasoning traces by final answer correctness to estimate the quality of intermediate latent states. During inference, reward gradients provide state-specific correction directions for fragile latent states, while a reward and confidence gate restricts intervention to states the reward signal flags as fragile. Experiments on multiple reasoning LLM backbones and benchmarks show that \ours consistently improves performance over various baselines, and post-hoc analyses further indicate that \ours implicitly promotes good cognitive behaviors that fix the original reasoning errors. Code is available at: https://github.com/jiakanglee/Latent-Reward-Steering. --- ### WaveFilter: Enhancing the Long-Context Capability of Diffusion LLMs via Wavelet-Guided KV Cache Filtering Authors: Jinnan Yang, Yan Wang, Zhen Bi, Kehao Wu, Xiaojie Li, Jungang Lou, Zechao Li, Jing Liu Published: 2026-05-30 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2606.00724v1 Diffusion Large Language Models (DLMs) have demonstrated significant advantages across various tasks. However, constrained by their multi-step iterative inference mechanism, their computational overhead and inference latency in long-context tasks have become core bottlenecks restricting their large-scale deployment. When processing long sequences, existing Key-Value (KV) caching mechanisms often face a dilemma where generation quality degrades drastically, where the core challenge lies in precisely and efficiently filtering critical tokens within ultra-long contexts. Inspired by the human reading process, we propose \textbf{WaveFilter}, a universal and training-free caching framework. This framework innovatively introduces the wavelet transform for decomposition of long sequences to achieve precise identification of key tokens, based on which a sparse KV Cache is constructed to compute the final contextual representation. Experimental results demonstrate that WaveFilter, as a plug-and-play generic framework, significantly enhances the performance of existing mainstream KV Cache methods in complex long-context tasks. --- ### EPIC: Efficient and Parallel Inference under CFG Constraints for Diffusion Language Models Authors: Hyundong Jin, Yo-Sub Han Published: 2026-05-30 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2606.00722v1 Controlling language model outputs is essential for ensuring structural validity, reliability, and downstream usability, and diffusion language models are no exception. Recent advances in diffusion language model decoding have extended output control beyond regular constraints to context-free grammar (CFG) constraints. Existing methods, however, can be up to four times slower than unconstrained decoding. More importantly, they substantially diminish one of the key advantages of diffusion language models over autoregressive models, namely parallel decoding. This slowdown arises because sequential validity checking introduces significant overhead during parallel generation. We propose an efficient CFG-constrained decoding framework, EPIC, that addresses this limitation. Our method improves decoding efficiency by combining lexing memoization, validation using Earley-style parsing instead of deterministic automata, and relaxed compatible subset selection for parallel commit. It reduces repeated lexing and validation overhead while allowing multiple compatible tokens to be committed together. Experiments on three benchmarks using four models show that our method reduces inference time by up to 67.5% and decreases the additional overhead by up to 90.5% compared with existing CFG-constrained decoding methods. Our implementation is available at https://github.com/hyundong98/EPIC-Decoding.git . --- ### LLM-Driven Co-Evolutionary Automated Heuristic Design for Bi-Component Coupled Combinatorial Optimization Authors: Mingen Kuang, Xudong Deng, Xi Lin, Ye Fan, Jianyong Sun, Jialong Shi Published: 2026-05-30 Categories: cs.AI, math.OC arXiv: https://arxiv.org/abs/2606.00718v1 While Large Language Models (LLMs) have recently shown promise in Automated Heuristic Design (AHD), existing methods typically generate and evolve heuristics as a single operator or search strategy, limiting their ability to model strong coupling among multiple decision substructures in problems such as the Traveling Thief Problem (TTP) and the Traveling Purchaser Problem (TPP). In this work, we propose CoEvo-AHD, an LLM-driven dual-population co-evolutionary framework for automated heuristic design in coupled combinatorial optimization. Unlike prior methods that evolve individual heuristics in isolation, CoEvo-AHD leverages LLMs to co-evolve two closely related operator populations. A cooperative evaluation mechanism explicitly captures interactions between route and selection operators, while pairwise scoring and synergistic joint crossover help discover complementary operator logic for joint improvement across coupled decision subspaces. We further design a tool-invocation environment library that encapsulates frequently used core operations, such as local-search delta computation, into callable functions, enabling LLM-generated operators to use standardized interfaces instead of reimplementing inefficient and error-prone problem-specific loops. Experiments on TTP and TPP show that CoEvo-AHD automatically discovers cooperative heuristic combinations and achieves competitive solution quality against traditional heuristics. --- ### Multi-Agent Conformal Prediction with Personalized Statistical Validity Authors: Martin V. Vejling, Christophe A. N. Biscio, Adrien Mazoyer, Petar Popovski, Shashi Raj Pandey Published: 2026-05-30 Categories: cs.LG, cs.AI, stat.ML arXiv: https://arxiv.org/abs/2606.00717v1 Uncertainty quantification is essential in high-stakes machine learning tasks. However, one of the principled solutions, conformal prediction, faces challenges under limited local calibration data, privacy constraints, and data heterogeneity. In multi-agent settings, existing works do not simultaneously and satisfactorily address these challenges with guarantees either limited to averages across agents or losing validity in heterogeneous settings. Hence, we propose personalized federated weighted conformal prediction (PFWCP), a framework that combines local density ratio weighting with weighted quantile aggregation to correct for heterogeneity while preserving privacy. The method yields asymptotically valid marginal and calibration-conditional coverage guarantees for each participating agent and supports protocols with one-shot communication. Theoretical analysis presents an adjustment to the coverage variance, governed by an effective sample size expression, which is necessary in the context of weighted conformal prediction, and experiments on synthetic and real datasets show improved calibration quality over state-of-the-art federated conformal baselines. --- ### Graph Transfer Learning via Shared Latent Geometry: Theory and Applications Authors: Tong Wu, Andrew Campbell, Anna Scaglione Published: 2026-05-30 Categories: cs.LG, eess.SP arXiv: https://arxiv.org/abs/2606.00716v1 Inference and control in engineered physical systems pay a heavy physics cost at deployment: state estimators, inverse-problem solvers, model-predictive controllers, schedulers, and observers are often not closed-form and must re-solve a numerical optimization per instance, with the operator re-supplied each time. Physics-informed learning moves this cost to training, but uses a single encoder pathway whose latent geometry de-learns under fine-tuning and admits no quantitative transfer guarantee. We propose an asymmetric two-pathway architecture that resolves both issues. A teacher encoder consumes privileged dense states from a high-fidelity simulator and represents the system through operator-polynomial features stable under spectral perturbation; a student encoder learns the same latent geometry from sparse field data and operator descriptors. At deployment the teacher is discarded, and the frozen student runs in a single forward pass with a transfer certificate. The design connects to privileged-information learning, knowledge distillation, and cross-modal distillation, but targets cross-instance transfer rather than fixed-instance prediction: topology and operator may change, while the latent task does not. We establish sufficient and near-necessary transfer conditions via Wasserstein proximity between latent laws, yielding a zero-shot error bound, and develop a finite-sample certification protocol with active expansion when coverage is incomplete. The framework applies wherever a system admits an operator with reportable spectrum. On power-system estimation, it achieves zero-shot transfer to 100 unseen topologies, a 95% certificate pass rate, accuracy competitive with topology-aware Newton--Raphson, and sub-millisecond inference. These results suggest asymmetric pathways plus operator-anchored latent geometry provide a foundation for certified zero-shot inference and control. --- ### Rate-optimal neural boundary detection from unlabeled noisy images Authors: Kyeongho Kim, Ilsang Ohn Published: 2026-05-30 Categories: stat.ME, stat.ML arXiv: https://arxiv.org/abs/2606.00715v1 We study boundary detection for unlabeled noisy images from a statistical perspective. The aim is to recover an unknown object region from raw intensity observations without pixel-wise annotating labels or a parametric model for the intensity distributions. Motivated by robust Gibbs posterior approaches based on thresholded misclassification losses, we propose a continuous hinge-type surrogate loss for boundary detection. The proposed loss is amenable to gradient-based optimization and can be combined with deep neural networks to represent complex object boundaries. We prove that the proposed loss function is Fisher consistent under a mild separation assumption and obtain a calibration inequality linking excess surrogate risk to the symmetric difference error of the estimated region. Under a piecewise smooth boundary model, we prove that the resulting deep neural network estimator achieves the minimax-optimal boundary recovery rate, up to logarithmic factors. The piecewise smooth formulation accommodates boundaries with corners and kinks, thereby extending beyond globally smooth boundary models. Numerical experiments demonstrate that the proposed method accurately and stably recovers object boundaries across a range of noise levels and shape configurations, and compares favorably with existing unsupervised boundary detection methods. --- ### CASTLE2026 Team WDL Technical Report Authors: Zhengyang Li, Zhenglin Du, Yi Wen, Fang Liu, Shuo Li, Xu Liu Published: 2026-05-30 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.00712v1 The CASTLE Challenge @ EgoVis 2026 evaluates long-form egocentric video question answering over 600+ hours of multi-perspective recordings. Each four-choice question requires evidence from videos, transcripts, auxiliary photos, people, days, rooms, and temporal context. We propose an evidence-aware multimodal reasoning pipeline based on Qwen. Our system parses question hints, retrieves ASR chunks, attaches auxiliary images, samples candidate video frames, and routes questions into static visual, speech/text, temporal, and mixed types with specialized prompts. Multiple inference passes are aggregated by confidence-weighted voting and converted into the official Codabench format. In ablation, LoRA improves the score from 0.21 to 0.50, and more sampled frames further raise it to 0.58. Our final system ranks first in the CASTLE Challenge @ EgoVis 2026. --- ### MOSAIC: Modular Orchestration for Structured Agentic Intelligence and Composition Authors: Yifan Bao, Xinyu Xi, Xinyu Liu, Wen Ge, Lei Jiang, Kevin Zhang, Raad Khraishi, Yihao Ang, Anthony K. H. Tung, Lukasz Szpruch, Hao Ni Published: 2026-05-30 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.00708v2 Automated data science is a structured model-selection problem. A solution must choose data transformations, feature representations, architecture, training procedure, evaluation protocol, and refinement strategy for a task. AutoML systems automate parts of this process, but typically search within predefined pipeline, model, and hyperparameter spaces. LLM-based agents offer greater flexibility through retrieval, code generation, and execution feedback, yet their modelling decisions are often unstructured, difficult to verify, and hard to reuse. We introduce \textsc{MOSAIC} (Modular Orchestration for Structured Agentic Intelligence and Composition), a structured agentic framework for memory-grounded model selection and workflow construction. Given a task and dataset, \textsc{MOSAIC} builds a semantic task profile, retrieves prior cases and source-code modules, and constructs a blueprint: an intermediate representation specifying selected modelling components, composition, interface constraints, and execution requirements. This blueprint turns model selection into a staged, context-grounded search and grounds LLM-based code generation in retrieved evidence rather than unconstrained synthesis. Candidate models are validated by execution and refined using diagnostic feedback, training traces, task metrics, and a failure-aware reinforcement learning policy. We instantiate \textsc{MOSAIC} on financial time-series forecasting and generation, where models must satisfy predictive accuracy, distributional fidelity, execution reliability, and downstream financial criteria such as risk and tail behaviour. Experiments against AutoML and agentic baselines show that \textsc{MOSAIC} improves task performance, execution success, and decision traceability, demonstrating the value of treating automated data science as structured, reusable, and execution-grounded model selection. --- ### MOSAIC: Modular Orchestration for Structured Agentic Intelligence and Composition Authors: Yifan Bao, Xinyu Xi, Xinyu Liu, Wen Ge, Lei Jiang, Kevin Zhang, Raad Khraishi, Yihao Ang, Anthony K. H. Tung, Lukasz Szpruch, Hao Ni Published: 2026-05-30 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.00708v1 Automated data science is a structured model-selection problem. A solution must choose data transformations, feature representations, architecture, training procedure, evaluation protocol, and refinement strategy for a task. AutoML systems automate parts of this process, but typically search within predefined pipeline, model, and hyperparameter spaces. LLM-based agents offer greater flexibility through retrieval, code generation, and execution feedback, yet their modelling decisions are often unstructured, difficult to verify, and hard to reuse. We introduce \textsc{MOSAIC} (Modular Orchestration for Structured Agentic Intelligence and Composition), a structured agentic framework for memory-grounded model selection and workflow construction. Given a task and dataset, \textsc{MOSAIC} builds a semantic task profile, retrieves prior cases and source-code modules, and constructs a blueprint: an intermediate representation specifying selected modelling components, composition, interface constraints, and execution requirements. This blueprint turns model selection into a staged, context-grounded search and grounds LLM-based code generation in retrieved evidence rather than unconstrained synthesis. Candidate models are validated by execution and refined using diagnostic feedback, training traces, task metrics, and a failure-aware reinforcement learning policy. We instantiate \textsc{MOSAIC} on financial time-series forecasting and generation, where models must satisfy predictive accuracy, distributional fidelity, execution reliability, and downstream financial criteria such as risk and tail behaviour. Experiments against AutoML and agentic baselines show that \textsc{MOSAIC} improves task performance, execution success, and decision traceability, demonstrating the value of treating automated data science as structured, reusable, and execution-grounded model selection. --- ### CR-JEPA: Cross-Modal Joint-Embedding Predictive Learning for Remote Sensing Image Retrieval Authors: Md Aminur Hossain, Ayush V. Patel, Nitant Dube, Biplab Banerjee Published: 2026-05-30 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.00706v2 Cross-modal remote sensing image retrieval aims to retrieve semantically related scenes across heterogeneous sensing modalities. This remains challenging because paired observations may differ substantially in imaging physics, spatial resolution, spectral configuration, and visual appearance. Moreover, a single retrieval projection trained with one objective may be insufficient to jointly support cross-modal semantic alignment and same-modal neighbourhood preservation. We propose CR-JEPA, a Cross-modal Retrieval Joint-Embedding Predictive Architecture for dual-modality remote sensing retrieval. The model uses modality-specific stems, a shared transformer trunk, and JEPA-style predictive objectives to estimate masked latent target features within and across modalities. Inspired by LeJEPA, we apply Sketched Isotropic Gaussian Regularization to raw retrieval projections to stabilize embeddings and mitigate collapse. CR-JEPA further employs a decoupled-head design with a unified retrieval head for same-modal retrieval and a cross-modal retrieval head for cross-modal search. We evaluate CR-JEPA on BEN-14K, CBRSIR_VS, and DSRSID. On BEN-14K, CR-JEPA improves S1 to S2 retrieval from 61.23% to 75.82% and S2 to S1 retrieval from 63.73% to 75.40% over X-JEPA, while also achieving competitive same-modal retrieval with fewer parameters. --- ### VICR: Visual In-Context Restoration for Real-World Image Super-Resolution Authors: Qichang Zhang, Hailong Wang, Baiang Li, Linhao Wang, Rong Fu, Erkang Cheng, Simon James Fong Published: 2026-05-30 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.00704v1 Real-world image super-resolution (Real-ISR) requires balancing structural fidelity to degraded observations with realistic detail synthesis. However, existing generative Real-ISR methods often rely on entangled conditioning mechanisms, leading to structural drift or semantically inconsistent details. To address this issue, we propose Visual In-Context Restoration (VICR), a Diffusion Transformer (DiT)-based framework that formulates Real-ISR as image completion. Specifically, we introduce a decoupled visual prior injection mechanism that derives local and global cues from the low-quality (LQ) image: local cues help recover image structures and support high-frequency detail synthesis, while global cues guide overall generation and promote semantic consistency. For ambiguous regions under severe degradation, VICR employs an inference-time agent to refine semantic prompts using visual evidence from the LQ input while keeping model parameters fixed. Experiments show that VICR achieves state-of-the-art performance across multiple Real-ISR benchmarks with only 127M trainable parameters. ---