# Research Papers — 2026-05 · bundle page 1 of 326 > Full-text markdown of papers #1–50 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 [Next bundle →](https://aipod.dev/papers/archive/2026-05/page/2.md) ### AIREP: A Protocol for Per-Decision Evidence in AI Runtime Governance Authors: Ali Toygar Abak Published: 2026-05-31 Categories: cs.AI, cs.CR arXiv: https://arxiv.org/abs/2608.21363v2 Runtime-governance evidence often collapses materially different events into one audit record: a decision may be made, an instruction dispatched or received, an action may or may not execute, and a resulting state may or may not be observed. This paper presents AIREP, a vendor- and model-independent protocol for per-decision AI runtime evidence. Its current wire model separates evidence into four artifact families: Decision, Control, Execution, and Effect. Artifacts use closed core schemas, explicit identities and digests, declared scope limits, RFC 8785 canonical JSON, domain-separated SHA-256 hashing, and pure Ed25519 signatures. A three-level assurance model distinguishes structural/hash consistency (AIREP-Core), verifier-accepted authorship (AIREP-Authenticated), and independently anchored chain-head freshness and non-truncation relative to an accepted witness (AIREP-Witnessed); these classes do not establish event truth. A structured reconciler preserves failure, missing evidence, unevaluated prerequisites, and indeterminate outcomes as distinct states. The released beta includes a four-family first-party producer, Python and Node reference-verification paths, adversarial/lifecycle corpora, and reproducible validation. Post-release first-party Hermes and LightEval integration exercises preserve explicit evidence boundaries without claiming adoption or interoperability. Independent implementation evidence exists separately for a v0.1.2 producer and a v0.2 consumer/verifier; because they target different frozen versions, they do not establish same-version producer-to-consumer interoperability. AIREP remains experimental. --- ### Position: Behavioral Systems Require Behavioral Tests Authors: Manuel Cherep, Nikhil Singh, Pattie Maes Published: 2026-05-31 Categories: cs.AI arXiv: https://arxiv.org/abs/2608.18081v1 Artificial agentic systems increasingly operate as behavioral systems by interacting with dynamic environments, pursuing goals, and adapting over time. Yet, current evaluation methods largely focus on performance outcomes, not the underlying behavioral processes that produce them. This paper argues that AI agents must be evaluated like other behavioral systems: through systematic observation, perturbation, and interpretation of their actions. We draw on lessons from the behavioral sciences to motivate this position, and propose a research agenda focused on developing rigorous behavioral tests. These include methods for recovering decision strategies from action sequences, constructing environments that isolate behavioral differences, and probing emergent dynamics in multi-agent systems. Taken together, these directions offer a roadmap for developing a science of AI behavior. --- ### Large Language Models in Mental Health: A Systematic Review of Applications, Innovations, and Ethical Challenges Authors: Yisong Chen, Yifan Gao, Sijing Yu, Chuqing Zhao, Yang Lu Published: 2026-05-31 Categories: cs.AI arXiv: https://arxiv.org/abs/2608.18080v1 We present a review on the applications of large language models (LLMs) in health, e.g., social media analysis, clinical conversational agents, therapy support tools, prompt engineering, multimodal learning, and ethical considerations. We integrate findings from interdisciplinary studies utilizing diverse data sources such as social media posts, electronic medical records, and multimodal inputs to enable early detection of depression, suicide risk assessment, personalized therapy support, and psychoeducational content generation. Our review highlights advancements in LLM models and annotation strategies that enhance interpretability and clinical relevance, while we also emphasize the critical role of prompt engineering for domain adaptation. We also discuss emerging multimodal fusion techniques integrating text, speech, and sensor data for improved mental health diagnosis and monitoring. Finally, we address ongoing ethical, sociotechnical, and regulatory challenges, and advocate frameworks to ensure safe, equitable, and accountable deployment of LLMs in real-world mental health care. --- ### LoRA-Diffusion: Parameter-Efficient Fine-Tuning via Low-Rank Trajectory Decomposition Authors: Iman Khazrak, Narges Nejad, Mohammadhossein Homaei, Mostafa M. Rezaee, Robert C. Green Published: 2026-05-31 Categories: cs.CL arXiv: https://arxiv.org/abs/2608.12328v1 Parameter-efficient fine-tuning methods such as LoRA have transformed the adaptation of large autoregressive language models, enabling task-specific customization with substantially fewer trainable parameters. However, these methods have not been successfully extended to diffusion-based language models, which generate text through iterative denoising rather than sequential token prediction. We propose LoRA-Diffusion, a parameter-efficient fine-tuning approach that applies low-rank decomposition to the denoising trajectory instead of model weights. Unlike weight-based LoRA, which modifies individual transformation matrices, our method learns low-rank perturbations to the entire diffusion path from noise to output. We introduce trajectory-level low-rank adapters that modify each denoising step, step-adaptive rank allocation across diffusion phases, and compositional multi-task learning that allows merging task-specific modules at inference without retraining. On SST-2, QNLI, and MRPC, we report token-level denoising validation accuracy over five random seeds. LoRA-Diffusion achieves the highest mean performance on SST-2 and strong performance on QNLI and MRPC. Joint multi-task training further shows that LoRA-Diffusion achieves the highest token-level accuracy among the evaluated methods. The approach reduces per-task storage compared with full fine-tuning and establishes a parameter-efficient fine-tuning framework for diffusion language models. --- ### Comparative Analysis of Multilingual Pre-trained Models for Nepali Automatic Speech Recognition Authors: Suman Paudel, Sarbin Sayami Published: 2026-05-31 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2608.12327v1 Multilingual pretrained models nominally support Nepali, yet no controlled benchmark has compared them under a single fine-tuning protocol. We fine-tune six pretrained models (XLSR-53, IndicWav2Vec, MMS-1B, Whisper-Medium, Whisper-Large-v3-Turbo, and Conformer-Hi) spanning CTC self-supervised, autoregressive encoder-decoder, and hybrid Conformer-CTC architectures, on the OpenSLR SLR54 Nepali corpus (~165 hours) using identical preprocessing, splits, optimizer, and family-matched learning-rate schedules. We evaluate Word Error Rate (WER), Character Error Rate (CER), and Real-Time Factor (RTF) on three independent test sets (OpenSLR, FLEURS, Common Voice). Whisper-Large-v3-Turbo (14.76% WER) and IndicWav2Vec (14.89% WER) tie at the top despite a 9x parameter gap and 40x pretraining-data gap, providing direct empirical evidence that language-family proximity in pretraining can substitute for raw scale for in-domain Nepali. CTC decoders run up to 29x faster than autoregressive Whisper at the same accuracy, flipping the practical deployment preference toward CTC under any latency budget. Massively multilingual pretraining (MMS-1B) yields the smallest out-of-domain degradation on FLEURS (+12.55 pp), indicating that scale buys robustness rather than peak in-domain accuracy. The resulting benchmark provides the first standardized, multi-model, efficiency-aware reference numbers for Nepali ASR. --- ### Beyond the Hivemind: Escaping LLM Homogeneity via Meta-Persona Anchoring and Sequential Temperature Scaling Authors: Tairan Fu, Javier Conde, Carlos Arriaga, Gonzalo Martínez, Pedro Reviriego, Javier Coronado-Blázquez Published: 2026-05-31 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2608.02618v1 Recent studies have identified an ``Artificial Hivemind'' effect in Large Language Models (LLMs) causing models to converge on a narrow, homogenized consensus even for open questions. This semantic collapse limits the diversity of AI, resulting in high inter-response similarity ($\approx 0.80-0.90$) even under high-temperature sampling. In this paper, we propose a novel mitigation framework to increase diversity: Meta-Persona Anchoring combined with Filtered Temperature Scaling (FTS). Our approach utilizes a two-stage generation process: first, the model is prompted to self-select a unique, idiosyncratic persona to anchor its starting point; second, we apply a dual-stage sampling sieve, utilizing Top-$p$ filtering to preserve grammatical validity followed by extreme temperature scaling ($T \ge 4.0$) on the surviving candidates to explore the broadened probability distribution. We evaluate our method using the INFINITY-CHAT dataset on state-of-the-art open weight models under $\sim$20B parameters. Our results demonstrate a significant reduction in semantic convergence, with average pairwise cosine similarity dropping from ($\approx 0.85$) to ($\approx 0.65$). Our scheme achieves a majority of questions below the 0.7 threshold, effectively reducing the gap between artificial mode collapse and human-level typological diversity. We provide our implementation as an open-source framework to enable more diverse and creative AI deployments. --- ### Verification Without Distrust: Reframing User-Side Oversight as Routine Epistemic Governance in Everyday Human-Chatbot Interaction Authors: Aung Pyae Published: 2026-05-31 Categories: cs.HC, cs.AI arXiv: https://arxiv.org/abs/2607.24761v1 Research on human-AI interaction has long framed verification of system outputs as a trust-contingent behavior that better-calibrated trust should reduce. We test this assumption in everyday human-chatbot interaction through a mixed-methods survey of 153 frequent chatbot users. Contrary to the canonical prediction, we find no detectable association between trust and verification, with the result robust across sensitivity analyses. Three further user-side practices - refinement, correction, and approval before automated actions - are widely endorsed and positively associated with satisfaction. The data reveal a substantive distinction between evaluative oversight (trust-decoupled, weakly tied to satisfaction) and interventionist oversight (weakly trust-correlated, strongly tied to satisfaction). A medium-to-large satisfaction-control gap shows that effective task outcomes do not produce a felt sense of agency. Qualitative findings identify instrumental mental models, failure-mode-specific doubt, and demand for epistemic infrastructure. We reframe user-side oversight as routine epistemic governance compatible with trust, and derive four design directions for scaffolded oversight in conversational AI. --- ### Keyword Matters: Unveiling the Energy Sensitivity of On-Device LLM Prompting Authors: Ruiyi Tao, Xiaolong Tu, Haoxin Wang Published: 2026-05-31 Categories: cs.AI, cs.NI arXiv: https://arxiv.org/abs/2607.22568v1 Large Language Models (LLMs) are increasingly deployed on mobile and embedded devices to improve privacy and reduce network latency. Yet on-device inference faces a fundamental constraint: high energy consumption on battery-powered, resource-limited hardware. While model compression and runtime acceleration have been widely studied, the effect of \emph{prompt design} on energy efficiency remains underexplored. This paper presents an empirical study of the relationship between prompt wording and energy consumption for on-device LLMs. Using real power measurements collected on a smartphone, we quantify how linguistic features, particularly imperative keywords and instruction structure, affect decoding length and total energy. Our results show consistent energy differences across verbs and tasks, indicating that prompt engineering is a lightweight lever for improving energy efficiency. --- ### A Formal Kinetic Theory for Zeroth-Order Newton Dynamics:Stein-Corrected Hessian Estimation and Curvature--Variance Trade-offs Authors: Shihao Ji, Mingyu Li, Zihui Song Published: 2026-05-31 Categories: math.OC, cs.AI arXiv: https://arxiv.org/abs/2607.22567v2 Zeroth-order Newton-type methods are useful when gradients and Hessians are unavailable, but they behave quite differently from first-order gradient-free methods. We develop a kinetic framework for algorithms that estimate both gradient and Hessian from black-box function values. The naive random-direction Hessian estimator turns out to be biased even on quadratics; a Gaussian--Stein correction is needed to estimate the Hessian of the Gaussian-smoothed objective. Linearizing the inverse Hessian exposes two noise channels: gradient noise preconditioned by the inverse Hessian, and Hessian noise transmitted through an inverse-Hessian sandwich. Under a noisy oracle the second channel carries the second-difference factor $μ_H^{-4}$. A small-mass kinetic lift links the finite-step Newton update to an underdamped phase-space model; the overdamped spatial limit yields a Lyapunov bound that exposes the curvature--variance trade-off between step size, batch sizes, smoothing radii, and regularization. Numerical experiments confirm estimator identities, the gradient and Hessian variance laws, dimension scaling, inverse-perturbation accuracy, and optimization behavior under query-budget and regularization ablations. --- ### A Formal Kinetic Theory for Zeroth-Order Newton Dynamics:Stein-Corrected Hessian Estimation and Curvature--Variance Trade-offs Authors: Shihao Ji, Mingyu Li, Zihui Song Published: 2026-05-31 Categories: math.OC, cs.AI arXiv: https://arxiv.org/abs/2607.22567v1 Zeroth-order Newton-type methods are useful when gradients and Hessians are unavailable, but they behave quite differently from first-order gradient-free methods. We develop a kinetic framework for algorithms that estimate both gradient and Hessian from black-box function values. The naive random-direction Hessian estimator turns out to be biased even on quadratics; a Gaussian--Stein correction is needed to estimate the Hessian of the Gaussian-smoothed objective. Linearizing the inverse Hessian exposes two noise channels: gradient noise preconditioned by the inverse Hessian, and Hessian noise transmitted through an inverse-Hessian sandwich. Under a noisy oracle the second channel carries the second-difference factor $μ_H^{-4}$. A small-mass kinetic lift links the finite-step Newton update to an underdamped phase-space model; the overdamped spatial limit yields a Lyapunov bound that exposes the curvature--variance trade-off between step size, batch sizes, smoothing radii, and regularization. Numerical experiments confirm estimator identities, the gradient and Hessian variance laws, dimension scaling, inverse-perturbation accuracy, and optimization behavior under query-budget and regularization ablations. --- ### MedLoCoMo: A Long-Context Multi-Session Medical Dialogue Benchmark for Large Language Models Authors: Zeyu Zhang, Ziqing Wang, Kaize Ding Published: 2026-05-31 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.22566v1 MedLoCoMo is a Medical Long-Context Memory benchmark for patient-specific clinical reasoning over multi-admission medical dialogue. Existing medical QA benchmarks largely test short context knowledge or single document grounding, leaving open whether LLMs can use, connect, and abstain over longitudinal patient histories. We build MedLoCoMo from deidentified MIMIC-IV and MIMIC-IV-Note records by constructing admission-level clinical packets, synthesizing grounded doctor-patient conversations, and generating evidence linked QA items over single-admission, cross-admission, and adversarial unanswerable settings. The benchmark contains 100 patient timelines averaging 1,669.8 turns, 29.7 sessions, and 74,512.2 tokens per conversation. Across the evaluated baselines, cross-admission reasoning is consistently harder than localized evidence use, even when models have long context windows or use external memory or retrieval methods. The code and MedLoCoMo benchmark release is available at https://github.com/leozzy13/MedLoCoMo for use and reproducibility. --- ### LeafData: An Agentic System for Data Migration Authors: Sadanand Katukuri, Rajasekhar Bada, Navya Induri, Rohit Gandham, Lynette Pinto, Joses Selvan, Abishek Krishnamoorthy, Joseph Rozario, Pu Tian, Pavan Poudel, Yalong Wu Published: 2026-05-31 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.21618v1 Modern data migration relies on JSON configuration to define data connection, pipeline logic, and orchestration behavior. This requires domain knowledge from users and is time-consuming and error-prone. In this paper, we present LeafData, an agentic system that converts user intent into validated and executable JSON configuration for data migration. Specifically, LeafData comprises a frontend chatbot and the backend service. The chatbot incrementally collects required information from users and performs schema-driven validation, while the backend service processes validated inputs and generates JSON configuration artifacts. These artifacts are directly consumable by orchestration platforms, enabling end-to-end pipeline generation and execution without manual coding. LeafData supports heterogeneous data migration across various data sources and connectors including relational databases, file-based systems, document-oriented databases, and REST APIs. --- ### Cross-Modal Hierarchical Fusion for from Multi-Sensor Ground Observation Authors: Xinze Zhang Published: 2026-05-31 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.30647v1 Dense volumetric reconstruction of cloud microphysical fields from sparse ground-based instruments remains an open problem, largely because the available measurements are heterogeneous in both modality and spatial coverage. We present AtmoFuseNet, a framework that fuses multi-view sky camera imagery with millimeter-wave cloud radar and ceilometer observations to produce 4D (three spatial dimensions plus time) estimates of cloud state and wind. The method operates in three stages: a cross-modal hierarchical aggregation module that combines image feature pyramids with instrument-derived vertical profiles through layer-wise cross-attention; a conditional variational refinement module that maps the resulting volume to physically consistent microphysical fields under differentiable radar and image forward models; and a correlation-based motion estimator that recovers per-voxel 3D wind vectors from consecutive volumetric reconstructions. On collocated observations from a semi-arid site, AtmoFuseNet reaches 0.026 g m^-3 liquid water content MAE and 1.18 m s^-1 wind speed MAE, improving over existing retrieval baselines. Ablation experiments isolate the contribution of each module. --- ### Multi-Agent DRL for QoS and Energy Optimization in RIS-Enabled Open-RAN Industrial 6G TN/NTN Networks Authors: Marwan Dhuheir, Thang X. Vu, Symeon Chatzinotas Published: 2026-05-31 Categories: cs.NI, cs.AI arXiv: https://arxiv.org/abs/2606.28339v1 Industrial 6G networks require ultra-reliable, low-latency, and energy-efficient connectivity in dynamic and blockage-prone environments, where conventional terrestrial deployments often fail to ensure stable coverage. Hence, in this paper, we propose a RIS-enabled Open-RAN framework for integrated terrestrial/non-terrestrial (TN/NTN) industrial 6G networks, in which UAVs-mounted reconfigurable intelligent surfaces (RISs) cooperate with ground radio units and a high-altitude platform (HAP) to enhance connectivity for dense industrial IoT devices. Owing to the high dimensionality and strong coupling among decision variables, conventional optimization techniques become computationally intractable. To overcome this limitation, the joint optimization problem of data rates, latency, and energy consumptions is formulated as a decentralized partially observable Markov decision process (Dec-POMDP) and solved using a multi-agent deep reinforcement learning framework. Simulation results show improvements of up to 75\% in data rate, 25\% latency reduction, and 16\% energy savings compared with state-of-the-art learning-based and non-RIS baselines, demonstrating the effectiveness of RIS-assisted Open-RAN intelligence for industrial 6G networks. --- ### Geometric Fairness-Aware Routing for Federated Edge Networks Authors: Ratun Rahman Published: 2026-05-31 Categories: cs.NI, cs.AI arXiv: https://arxiv.org/abs/2606.26125v1 Emerging 6G and edge-intelligent networks require effective and balanced routing algorithms among varied and spatially distributed devices. Existing federated routing systems often prioritize aggregate latency or throughput above fairness and the underlying geometric structure of network topologies. This paper describes Geo-FairFed, a geometric fairness-aware routing system that blends hyperbolic graph neural networks (HGNNs) and federated optimization to provide equal performance across edge nodes. Each node learns topology-aware representations on a negatively curved manifold, which include hierarchical relationships and connection asymmetries. A global aggregator next enforces fairness using a curvature-regularized aim that minimizes routing loss, geometric inconsistency, and an inequality penalty based on Jain's fairness index. A theoretical analysis develops convergence guarantees under limited curvature and shows that the proposed fairness term results in a Pareto-improving equilibrium in routing performance. Extensive simulations on dynamic 6G-edge and IoT topologies reveal that Geo-FairFed minimizes average latency by 20\%, reduces energy consumption by 17\%, and improves fairness by up to 21\% when compared to state-of-the-art federated and geometric routing protocols. The study found that embedding topology in a hyperbolic manifold and including fairness into federated updates can significantly enhance the efficiency and equity of large-scale network routing. --- ### Measuring the Occupation-Level Impact of AbbVie Intelligence: AI Applicability Analysis, 2024-2025 Authors: John Regan, Jon Stevens, Brian Martin Published: 2026-05-31 Categories: cs.CY, cs.AI arXiv: https://arxiv.org/abs/2606.20635v1 This paper presents an empirical analysis of AbbVie Intelligence's measurable impact on employee work activities across 192 distinct occupations in 2024 and 2025. Drawing on 598,744 de-identified AI conversations classified according to the O*NET Intermediate Work Activity (IWA) taxonomy, we compute occupation-level AI Applicability Scores that quantify the extent to which AI tools can meaningfully assist or automate real work at scale. Three convergent analyses are conducted: (1) longitudinal year-over-year trends from 2024 to 2025, (2) a quasi-experimental pre-post evaluation of the AbbVie Intelligence version 3 platform release in August 2025, and (3) a pre-post evaluation of the AbbVie AI Learning Summit held in November 22025. Results demonstrate statistically significant improvements across all three dimensions. Mean AI Applicability Scores rose substantially from 2024 to 2025; the platform release product a +10.0% gain (p<0.001); and the AI Learning Summit produced a +6.68% gain (p<0.001). These findings establish that both technological platform enhancements and structured enterprise AI eduction programs independently and substantially expand the reach of AI across the AbbVie workforce. --- ### Interpolation between Convolution and Attention via K-Nearest Neighbors Authors: Mingi Kang Published: 2026-05-31 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.14725v1 The shift from Convolutional Neural Networks to Transformers has reshaped computer vision, yet these two architectural families are typically viewed as fundamentally distinct. Convolutional Neural Networks are defined by spatially local convolution operations, while Transformers rely on global self-attention. We argue that convolution and self-attention, despite their apparent differences, can be unified within a single k-nearest neighbor aggregation framework. The critical insight is that both operations are special cases of neighbor selection and weighted aggregation. Convolution selects neighbors by spatial proximity while self-attention selects by feature similarity, revealing that they lie on a continuous spectrum rather than representing categorically different computations. We introduce Convolutional Nearest Neighbors (ConvNN), a unified framework that formalizes this connection. ConvNN exactly recovers standard and depthwise convolution by restricting neighbor selection to normalized spatial coordinates, and exactly recovers self-attention and its sparse variants, including KVT-attention, by replacing spatial proximity with scaled dot-product similarity. Beyond these special cases, ConvNN serves as a drop-in replacement for both convolution and attention layers, enabling systematic exploration of the intermediate spectrum between local and global aggregation through configurable similarity functions, neighbor selection strategies, positional encodings, and aggregation kernels. --- ### VigilFormer: Deformable Attention for Video Anomaly Detection with Causal Risk Inference Authors: Xinze Zhang Published: 2026-05-31 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.14724v1 Video anomaly detection in surveillance settings must balance detection accuracy against real-time throughput, a tension that existing methods address either through stronger feature extractors or more efficient architectures, but rarely both. We present VigilFormer, a unified framework that combines deformable spatio-temporal attention with causal temporal modeling to detect anomalies in untrimmed surveillance video. The proposed Deformable Spatio-Temporal Encoder (DSTE) attends to a sparse set of informative locations across frames, avoiding the quadratic cost of dense attention while retaining the ability to capture irregular motion patterns. A Causal Anomaly Classifier (CAC) applies dilated causal convolutions over snippet-level features and optimizes a contrastive multiple-instance learning objective that separates anomalous and normal representations without frame-level labels. To meet deployment constraints, an Adaptive Confidence Scheduler (ACS) dynamically skips low-information frames at inference time, reducing redundant computation in static scenes. Evaluated on UCF-Crime, ShanghaiTech, and CUHK Avenue, VigilFormer achieves AUC scores of 87.83%, 97.21%, and 89.74% respectively, at 41.5 FPS on a single GPU, outperforming recent weakly-supervised methods in both accuracy and speed. --- ### Disagreement-Based Cross-Model Routing for Implicit Video Question Answering Authors: Durga Sandeep Saluru Published: 2026-05-31 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.14723v1 We study multiple-choice video question answering on the ImplicitQA benchmark, where the correct answer is never explicitly shown but must be inferred from off-screen events, line-of-sight cues, causal structure, and cross-shot spatial layout. On this benchmark a single frontier video LLM already operates near its accuracy ceiling, and we observe that conventional self-consistency strategies -- majority voting across repeated samples of the same model -- can hurt rather than help, because the model's errors on hard questions are correlated. We propose disagreement-based cross-model routing, a pure inference-time procedure that requires no labels and no training. We triple-sample a native-video model (Gemini 3.1 Pro Preview) at temperature zero, exploit the genuine sample-to-sample variance of its video-processing pipeline to identify the roughly 20% subset of questions where the three samples disagree, and route only that subset to a second model from a different family (Claude Opus 4.8) that consumes uniformly sampled frames with adaptive thinking. On the 1001-question validation set with public ground truth -- our main evaluation -- the method improves AvgAcc by +1.43 over the best single sample of the primary model, with per-category gains concentrated on Motion & Trajectory (+5.49), Inferred Counting (+3.45), and Vertical Spatial Reasoning (+1.82) -- the categories most dependent on cross-shot reference resolution. The same pipeline applied to the held-out 172-question CVPR 2026 ImplicitQA challenge test set achieves 82.03 AvgAcc / 79.71 MacroAvgAcc (+1.81 over the best single sample of the primary model), confirming the validation result on an independent split. --- ### Active Inference for Adaptive Traffic Signal Control in Noisy Nonstationary IoT Environments Authors: Dénes Toth, George Ambroladze, Edwin Sundberg, Ali Beikmohammadi, Alfreds Lapkovskis Published: 2026-05-31 Categories: eess.SY, cs.AI, cs.LG, cs.NI, cs.PF arXiv: https://arxiv.org/abs/2606.13698v1 Urban traffic signal control at IoT-instrumented intersections must remain effective under sensor occlusion, weather attenuation, and nonstationary demand. Conventional controllers degrade under these conditions, and learned policies remain difficult to audit. To address these challenges, we propose an active inference controller for a four-arm signalized intersection that dynamically selects phases by minimizing expected free energy (EFE) over Gaussian beliefs about per-direction congestion levels, yielding a fully traceable decision pipeline. We benchmark the controller in a SUMO traffic simulator against a rule-based heuristic and a deep Q-network (DQN) across four scenarios that progressively increase noise and nonstationarity, spanning sensor occlusion, adverse weather, and stochastic accidents. Across 100 independent random evaluations per scenario, active inference attains the lowest idle times and CO2 emissions in the noisiest scenarios (56,977 s and 29.12 kg vs. 71,741 s and 30.56 kg for DQN). These gains come at a modest cost in bus priority service rate and phase switch frequency. --- ### AGORA: Can Deliberation and Governance Gates Absorb Participation Bias in Transit Planning? Authors: Jung-Hoon Cho, Cathy Wu Published: 2026-05-31 Categories: cs.CY, cs.LG, cs.MA, cs.SI arXiv: https://arxiv.org/abs/2606.13696v1 Transit network design depends not only on the optimization algorithm but also on who shows up to the public hearing. Current practice often collects one-directional comments from self-selected attendees, leaving participant mix as an uncontrolled source of outcome variation. We present AGORA, a framework that holds the network, demand, and solver fixed while systematically varying meeting composition through stakeholder agents, structured deliberation, and governance gates. Across two standard benchmark networks at different scales, we find that (i) aggregate outcomes vary little across compositions, but on tail risk and fairness disparity, representative sampling still tends to outperform skewed compositions; (ii) without deliberation, composition produces no variation at all, showing that deliberation is the mechanism through which who attends affects outcomes; and (iii) governance gates compress cross-profile variance without shifting the average outcome on Mandl, but low acceptance on Mumford0 shows thresholds require instance-specific calibration. These findings reframe participation bias from an uncontrollable input to a process-design problem: even without guaranteed representative attendance, well-structured deliberation and governance criteria can substantially reduce how much outcomes depend on who is in the room. --- ### Korzhinskii-Net: Physics-Informed Neural Network for Sub-Surface Mineral Prospectivity Modelling Authors: Boris Kriuk Published: 2026-05-31 Categories: physics.geo-ph, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.13695v2 Mineral prospectivity modelling (MPM) underpins exploration economics, yet most operational pipelines reduce to data-driven classifiers trained on shallow surface proxies. Such models are blind to the subsurface physics that actually localises ore: heat advection, fluid flow, and lithology-dependent precipitation. We present Korzhinskii-Net, a 2-D radial physics-informed neural network (PINN) that couples Darcy flow, advective-diffusive heat transport, and a softplus-saturated reaction rate into a single differentiable forward model, weakly supervised by surface and remote-sensing proxies. The network is named after Dmitri S. Korzhinskii (1899-1985), whose theory of infiltration metasomatism provides the physical scaffold. We evaluate Korzhinskii-Net on six ore provinces spanning three commodity classes - Udokan (sandstone-hosted Cu), Sukhoi Log, Olimpiada, and Berezovskoye (orogenic Au), Vorontsovskoye (Carlin-type Au), and Dalnegorsk (skarn polymetallic) - under a fair, leakage-controlled 5-fold cross-validation protocol with hard ring-shaped negatives and baseline proxy features disabled. Korzhinskii-Net attains a mean PR-AUC of 0.708 versus 0.235 for the strongest classical baseline (support vector machine), and a mean fractional rank of 0.036 versus 0.475. The improvement is consistent across all six provinces and three commodity systems, suggesting that physics-informed differentiable simulators, even when constrained only by global open-data proxies, can recover localisation patterns that pure feature-based learners systematically miss. We release the full pipeline and evaluation harness as open source. --- ### Efficient Temporal Modeling for Mobile Sleep Staging via Lightweight Random Attention Authors: Guisong Liu, Pengfei Wei, Jainsong Zhang, Martin Dresler Published: 2026-05-31 Categories: eess.SP, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.13694v1 Mobile sleep staging serves as a foundational infrastructure for in-home sleep monitoring and closed-loop modulation. But existing sequential models such as RNNs and Transformers are computationally expensive for mobile deployment. In this paper, we propose Random Attention (RA), a lightweight temporal modeling module based on fixed random projections, which replaces learnable sequence modeling with similarity-based aggregation. RA introduces little additional parameters beyond the epoch encoder while enabling effective temporal smoothing. We further provide a theoretical interpretation via the Random Attention Prior Kernel (RAPK), which decomposes RA into a global smoothing term and a feature similarity term, offering an interpretable view of temporal sleep structure. Experiments on Sleep-EDF-20 and Sleep-EDF-78 show that RA consistently improves epoch-wise baselines by 1-3\% in accuracy and F1 score, while achieving competitive performance compared with LSTM, GRU, and Transformer models. RA also demonstrates strong generalization across different backbone encoders and improved robustness over conventional temporal smoothing methods. These results indicate that efficient sleep staging can be achieved through lightweight similarity-based temporal aggregation, making RA suitable for real-time wearable applications. --- ### Blurry Window Attention Authors: Axel Laborieux, Christos Sourmpis, Juan Gabriel Kostelec, Qinghai Guo Published: 2026-05-31 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2606.09862v1 The Softmax Attention operation in Transformer language models has a quadratic complexity in the sequence length and a growing state size in the form of KV cache, which becomes a bottleneck in long context scenarios. To overcome this limitation, alternative architectures with linear complexity and finite state size have been introduced, such as State-Space Models (SSMs), Linear Attention (LA), and Attention with Bounded-memory Control (ABC). Though linear models achieve similar language perplexity as Transformers, they are still behind in tasks which require retrieval or recall of specific information. In this work, we introduce Blurry Window Attention (BLA) a novel ABC method inspired by SSMs. BLA stores a frequency window from which a blurry KV history is reconstructed via interpolation using Dirichlet kernels. BLA can be understood as a generalization of Sliding Window Attention (SWA) depending on the Dirichlet kernels resolution or as a special case of the Gated Slot Attention (GSA), where the decay factor is implemented with Dirichlet kernels. We describe in details the theory and efficient implementation of BLA. On the Multi-Query Associate Recall (MQAR) synthetic task, we show that the state efficiency of BLA is 8$\times$ better than SWA and is competitive with popular linear attention models, and in the RegBench synthetic task, only BLA and SWA improve their performance as the state size grows among the linear models we tested. --- ### Time Series as Language: A Universal Tokenizer for General-Purpose Time Series Foundation Models Authors: Yunhao Zhang, Ruiying Qi, Jiale Zheng, Jianfeng Zhang, Lujia Pan, Junchi Yan Published: 2026-05-31 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2606.09861v1 While Next-Token Prediction (NTP) has unified LLM pretraining, its adaptation to unbounded, continuous time series (TS) remains open. To bridge the gap, we introduce UniTok, a universal tokenizer that transforms TS into discrete tokens, and UniTok-FM, a foundation model pretrained via NTP on these tokens. UniTok-FM is a general-purpose foundation model that supports zero-shot and prompt-boosted forecasting, as well as few-shot generation and classification via training-free in-context inference--a capability not achieved by prior works. Technically, UniTok is a vector-quantized autoencoder incorporating prefix normalization for scale stabilization, a progressive-resolution causal architecture for encoding and decoding, and a structure-preserving reconstruction loss for training. UniTok-FM adopts an off-the-shelf LLM architecture without TS-specific modifications. Instead of pretraining on isolated TS, it performs NTP on context windows formed by multiple series with similar patterns, aiming to capture their shared dynamics. Experiments on forecasting, generation, and classification show that a single unified UniTok-FM consistently outperforms statistical and supervised baselines, achieves competitive performance with task-specific foundation models, and uniquely enables training-free in-context inference across tasks. --- ### Conformal Risk Prediction for Non-Alcoholic Fatty Liver Disease Using Gradient Boosting with Distribution-Free Coverages Authors: Xinze Zhang Published: 2026-05-31 Categories: cs.LG, cs.AI, stat.AP, stat.ML arXiv: https://arxiv.org/abs/2606.09860v1 Non-alcoholic fatty liver disease (NAFLD) affects roughly 25% of global adults, posing substantial hepatic and cardiovascular risks. Yet, population-level screening tools remain inadequate. We present Method, a machine-learning framework for NAFLD risk prediction coupling gradient-boosted decision trees with conformal prediction to yield calibrated, distribution-free coverage guarantees on individual risk estimates. It integrates a mutual-information-based stability selection procedure to identify a compact, clinically interpretable feature subset via bootstrap resampling, constructing prediction sets whose marginal coverage provably exceeds a user-specified confidence level. We evaluated Method on a multicenter cohort from Guangzhou, China (primary n=2,187; external validation n=412) using 78 candidate features across demographics, metabolic biomarkers, and lifestyle factors. Method achieves an AUROC of 0.912 internally and 0.891 externally, outperforming deep neural networks, TabNet, support vector machines, and logistic regression. Conformal prediction sets achieve 91.3% empirical coverage at the 90% nominal level. A three-tier risk stratification derived from these scores separates the population into distinct groups, with the high-risk subgroup showing a 12-month progression rate 4.7 times that of the low-risk tier. The selected features -- notably waist circumference, ALT, GGT, triglycerides, fasting glucose, and BMI -- align with established metabolic risk factors, providing biological plausibility. --- ### Mitigating Manifold Departure: Uncertainty-Aware Subspace Rectification for Trustworthy MLLM Decoding Authors: Yingxuan Zhuang, Jingxiao Yang, Miao Pan, Cheng Tan, Yuxiang Cai, Siwei Tan, Chen Zhi, Xuhong Zhang, Jianwei Yin, Jintao Chen Published: 2026-05-31 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2606.09859v1 MLLMs frequently hallucinate objects inconsistent with visual inputs. This issue is typically attributed to the over-reliance on language priors, which can override the visual context. Recent training-free decoding strategies address this by penalizing language priors. However, these methods overlook the dual nature of language priors, where they can be both helpful and harmful depending on the alignment with visual evidence. In particular, blindly suppressing language priors often disrupts the model's semantic manifold, leading to performance degradation, a phenomenon we term Manifold Departure. To address this, we propose Manifold-Guided Adaptive Projection (MGAP), a geometry-aware, training-free decoding method that mitigates hallucinations while preserving representation structure. MGAP first constructs a language-prior subspace from blind hidden states via SVD. During decoding, MGAP projects each multimodal hidden state onto this subspace and applies a consistency-aware gate to adaptively attenuate only the projected prior component, yielding a subspace-selective update that largely preserves the orthogonal semantic components. Extensive experiments on POPE and CHAIR show that MGAP outperforms prior decoding baselines, achieving stronger hallucination suppression without sacrificing coherence. --- ### Crayotter: Traceable Multi-Agent Workflows for Long-Form Video Editing Authors: Lecheng Yan, Yichong Zhang, Xiantao Xu, Jianze Lin, Ben Pan, Xiaoyu Zheng, Jiawei Qian, Anqi Wu, Jiahui Geng, Ruizhe Li, Fengyu Cai, Jingcheng Niu et al. Published: 2026-05-31 Categories: cs.CV, cs.CL, cs.MA arXiv: https://arxiv.org/abs/2606.07636v2 Long-form video editing over heterogeneous footage requires agents to coordinate source selection, multimodal analysis, timeline construction, narration and subtitle alignment, rendering, and revision while exposing intermediate state for inspection and repair. We present Crayotter, an open-source multimodal multi-agent demo system for prompt-driven long-form video editing. Crayotter organizes production around coverage-aware material preparation, artifact-grounded editing research, and tool-grounded timeline execution. Across these stages, retrieval reports, video analyses, editing blueprints, scheduler events, tool calls, intermediate renders, and final exports are treated as first-class artifacts rather than hidden transient state. The workbench supports local assets, agent-assisted retrieval, progress monitoring, artifact preview, failure diagnosis, interrupted-job resumption, and resource-aware asynchronous execution for long-running workflows. In a 23-theme evaluation, Crayotter achieves the highest human overall score (3.40/5) among the compared systems, with its largest margins in theme alignment, narrative coherence, and editing smoothness. These results show that long-horizon video editing agents can be made traceable, inspectable, and practically controllable through observable production artifacts. Code, traces, and examples are publicly available at https://github.com/idwts/Crayotter. --- ### NeuroAlign: Hierarchical Multimodal Fusion of Dynamic and Structural Neuroimaging for MCI Analysis Authors: Xiongri Shen, Zhenxi Song, Jiaqi wang, Yi Zhong, Leilei Zhao, Chenqi Xu, Linling Li, Yichen Wei, Lingyan Liang, Demao Deng, Luping Song, Ping Luan et al. Published: 2026-05-31 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.07635v1 Multimodal neuroimaging fusion of functional MRI (fMRI) and diffusion tensor imaging (DTI) provides complementary information for cognitive impairment analysis, but remains challenged by heterogeneous feature spaces and misaligned representations. We propose \textit{NeuroAlign}, a hierarchical framework for structured multimodal fusion. It introduces (1) \textit{Dual-Modal Hierarchical Alignment} (DMHA), which models multi-scale dynamic connectivity and aligns dynamic-static and functional-structural embeddings; and (2) \textit{Dual-Domain Hierarchical Interaction} (DDHI), which enables fine-grained modulation and global interaction between connectivity- and region-level features. To support feature-level inspection, we design \textit{Synergistic Activation Mapping} (SAM), a gradient-free, marker-oriented attribution method for DFC, SFC, ALFF, and FA. Evaluated on GUTCM, ADNI, and OASIS under five-fold validation, NeuroAlign achieves competitive MCI/SCD detection and preliminary cross-dataset transferability. Attribution analyses reveal modality-specific and partially consistent brain patterns, providing model-derived evidence for multimodal representation analysis. --- ### AMN: An Adaptive Multi-Scale Fusion Network with Boundary and Uncertainty Modeling for Nuclei Segmentation Authors: Spoorthi M, Suja Palaniswamy Published: 2026-05-31 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.07633v1 Accurate classification of nuclei subtypes in histopathology images is critical for downstream tasks including tumor grading, immune infiltrate quantification, and prognosis prediction. Existing approaches rely on either convolutional or transformer-based encoders in isolation, limiting their ability to simultaneously capture fine-grained local texture and long-range spatial context. We present AMN (Adaptive Multi-Scale Nuclei Network), a dual-encoder segmentation framework that jointly leverages a Swin Transformer and a ResNet-50 feature pyramid, fused via a learned per-channel gating mechanism that dynamically weighs each encoder's contribution at every scale. AMN is trained with a multi-objective loss combining class-weighted focal loss, boundary-aware loss with positive-pixel emphasis, and a novel uncertainty-modulated classification term that suppresses overconfident erroneous predictions. Evaluated on the CoNIC benchmark across seven nuclei classes, AMN achieves a mean Dice of 0.82 and mean F1 of 0.68, with an F1 of 0.67 on the diagnostically challenging lymphocyte class. AMN outperforms eight baseline models spanning pure-CNN, pure-transformer, and recent hybrid architectures: U-Net, ResU-Net, DeepLabV3+, SegNet, ViT-Small, HmsU-Net, ConvFormer-UNet, and BEFUnet. Cross-dataset evaluation on MoNuSeg demonstrates strong generalization without retraining and validating the domain robustness of the learned representations. --- ### Evaluation of ML Resource Utilization Requires Model Life Cycle Assessment Authors: Jared Fernandez, Clara Na, Yonatan Bisk, Constantine Samaras, Emma Strubell Published: 2026-05-31 Categories: cs.LG arXiv: https://arxiv.org/abs/2606.07632v1 Proper accounting of the energy requirements and environmental impact of artificial intelligence (AI) systems is necessary for researchers, developers, policy makers, and users to assess the barriers to building systems at scale. With the growing complexity of pipelines and underlying infrastructure needed to develop and deploy AI systems, previous approaches for evaluating AI efficiency which focus on the costs of a single training run or an individual inference prediction are no longer sufficient. In this position paper, we enunciate the need for applying life cycle assessment to evaluate the costs of the machine learning model development and deployment pipeline to properly account for the required resources and downstream impact. Life cycle assessments enable the incorporation of costs across the full life cycle of an AI system and its underlying infrastructure, from the embodied costs associated with the physical computing hardware through the operational costs in training and inference. --- ### Trait-space Monitoring for Emergent Misalignment During Supervised Finetuning Authors: Huy Nghiem, Sy-Tuyen Ho, Sarah Wiegreffe, Hal Daumé Published: 2026-05-31 Categories: cs.LG, cs.AI, cs.CY arXiv: https://arxiv.org/abs/2606.07631v1 Emergent misalignment (EM) occurs when narrow finetuning causes a model to behave dangerously outside the finetuning task. Standard training signals can miss this shift, making reliable detection costly if it depends on repeated behavioral evaluation. We ask whether emergent misalignment can instead be detected from internal representations during finetuning. Using seven alignment-relevant traits encoded as linear directions in activation space, we track representational drift across training checkpoints in four open-source 7-9B LLMs. EM-relevant drift concentrates on a low-dimensional axis that explains 65.5% of the variance, revealing a geometric signature in the studied regime. A low-overhead monitor built on this drift profile detects dangerous checkpoints with 2.2% false negative rate, 2.9% false positive rate, and 0.990 AUROC on held-out perturbation types, outperforming unsupervised PCA and SAE baselines. Stress tests on two 14B models, longer finetuning runs, and misaligned starting points identify key deployment boundaries. These results position trait-space monitoring as a practical complement to behavioral evaluation for EM detection during LoRA-based finetuning, while showing that deployment across substantially different regimes may require recalibration. --- ### The Ringelmann Effect in Multi-Agent LLM Systems: A Scaling Law for Effective Team Size Authors: Blaž Bertalanič, Carolina Fortuna Published: 2026-05-31 Categories: physics.soc-ph, cs.AI, cs.MA arXiv: https://arxiv.org/abs/2606.02646v1 Inference-time multi-agent LLM scaling lacks a shared unit: counting nominal agents conflates cost with independent evidence. We derive a two-parameter scaling law $R(N) = N_\text{eff}/N = 1/(1+c(N-1)N^{-β})$ where the regime exponent $β$ classifies any configuration into one of three asymptotic regimes -- hard-ceiling at $1/c$ ($β= 0$), sublinear at $N^β/c$ ($0 < β< 1$), or linear ($β\ge 1$), and a mean-field theorem predicts that peer count $k$ and rounds $τ$ during agent debate enter the dynamics only through their product $kτ$. The law applies at two levels: answer diversity and correctness redundancy. Across 44 (model $\times$ task $\times$ condition) cells spanning peer debate, self-correction, random-noise placebo, self-consistency, three open-weight families (Qwen, Llama, Ministral) at scales from 7B to 32B with a frontier API check (Gemini), thinking models, heterogeneous teams, and sparse communication, the functional form fits every condition at $R^2 > 0.99$; only $(c, β)$ shifts. On free-form math, dense peer influence collapses the answer-level regime from sublinear into hard-ceiling; correctness-level fits remain hard-ceiling throughout. Three findings have practical implications. \emph{(i)}~Thirty dense debating agents produce no more answer diversity than one on MMLU-Hard. \emph{(ii)}~A noise placebo tracks self-correction on free-form math and at $4\times$ scale, so within homogeneous teams the gain commonly attributed to ``debate'' comes from re-evaluation, not peer content. \emph{(iii)}~A single $N \le 5$ pilot predicts the $N=30$ structural ceiling, and within the configurations tested only architectural diversity (heterogeneous teams) lowers $c$ and escapes the hard-ceiling regime, communication-mode interventions do not. --- ### Target Updates May Stabilize Linear Q-Learning: Periodic and Soft Dynamics Authors: Donghwan Lee Published: 2026-05-31 Categories: stat.ML, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.02645v1 Periodic target updates in Q-learning and soft target updates in actor-critic methods are empirically well established stabilization mechanisms, but their precise theoretical explanation is still incomplete. This paper gives a rigorous and exact analysis of these mechanisms for Q-learning with linear function approximation (linear Q-learning) using the exact switched linear system (SLS) dynamics induced by the Bellman maximum and the joint spectral radius (JSR) of the resulting switching matrix families. Although linear Q-learning can fail to converge in general, we prove that, under explicit spectral and step-size conditions, periodic hard target updates and soft target updates can guarantee convergence to the exact projected Q-Bellman solution. The main analysis is carried out for deterministic linear Q-learning, where the target-update mechanism is most transparent. Once the corresponding JSR certificate is established for the mean recursion, the stochastic reinforcement-learning setting can be treated by replacing deterministic modes with sampled stochastic modes and adding the corresponding stochastic-noise analysis. --- ### A New Framework for Cybersecurity Refusals in AI Agents Authors: Eliot Krzysztof Jones, Mateusz Dziemian, Matt Fredrikson, J Zico Kolter Published: 2026-05-31 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2606.02644v1 Agentic scaffolds have dramatically improved LLM performance on complex, long-horizon tasks, yielding both broad benefits and amplified risks in domains like cybersecurity. Existing benchmarks for AI agents in cybersecurity focus mainly on measuring proficiency--how effectively agents can complete offensive security tasks--but neglect a critical question: when and how should agents refuse harmful requests? We present the first framework for establishing refusal boundaries in offensive security contexts. Our framework defines (1) principled criteria for when tasks should be refused, (2) categories of tasks that warrant refusal, and (3) evaluation methodology for measuring agent robustness under both benign and adversarial conditions. We apply this framework to assess how current LLM-powered agents adhere to appropriate refusal boundaries across a range of web-based offensive security scenarios, finding that 6 of 8 frontier models tested show near-zero refusal rates, with only 2 models (GPT-5.2 and GPT-5.1 Codex) demonstrating any meaningful refusal behavior. --- ### Inference Cost Attacks for Retrieval-Augmented Large Language Models Authors: Chengliang Liu, Liangbo Ning, Yujuan Ding, Wenqi Fan Published: 2026-05-31 Categories: cs.CR, cs.AI, cs.DB arXiv: https://arxiv.org/abs/2606.02643v1 Retrieval-Augmented Generation (RAG)-enhanced LLM systems, while powerful, introduce substantial inference costs due to the inclusion of an extra multi-stage pipeline that dynamically retrieves and synthesizes information from external knowledge sources. This high operational cost exposes a critical vulnerability to Inference Cost Attacks (ICAs). However, existing ICAs often rely on the impractical assumption of direct prompt manipulation. We argue that a more feasible and potent threat to RAG-enhanced LLM systems arises from poisoning external knowledge bases (e.g., web knowledge from the Internet). In this work, we introduce the Retrieval-Augmented Inference Cost Attack (RA-ICA), a novel attacking paradigm that targets the computational cost of RAG-enhanced LLM systems by injecting malicious documents into external knowledge corpus. To operationalize this attack, we propose Computational Resource Exhaustion via External Poisoning (CREEP), a novel framework that leverages LLM agents to automatically craft malicious documents that are both semantically relevant for retrieval and potent for inducing an abnormal increase in token consumption during the inference phase. To enhance the attack's effectiveness, we introduce Memory-Augmented Group Relative Policy Optimization (MA-GRPO), a novel reinforcement learning algorithm that fine-tunes the agents by learning from a dynamic memory of historical best adversarial documents. Extensive experiments across three real-world datasets demonstrate that RA-ICA increases token consumption by up to 13.12 times with an over 90% success rate, without degrading the integrity of the generated answer. --- ### SVHalluc: Benchmarking Speech-Vision Hallucination in Audio-Visual Large Language Models Authors: Chenshuang Zhang, Kyeong Seon Kim, Chengxin Liu, Tae-Hyun Oh Published: 2026-05-31 Categories: eess.AS, cs.AI, cs.CV, cs.LG, cs.MM, cs.SD arXiv: https://arxiv.org/abs/2606.02642v1 Despite the success of audio-visual large-language models (LLMs), they can produce plausible but ungrounded outputs, termed hallucination. Existing benchmarks focus on environmental sounds (e.g., dog barking) to indicate event occurrence. In contrast, human speech carries fundamentally different, rich semantics and temporal structures, yet it remains unexplored whether current models can accurately align speech content with corresponding visual signals. In this work, we show that speech content can induce hallucinations in audio-visual LLMs. To systematically study this, we introduce SVHalluc, the first comprehensive benchmark for evaluating speech-vision hallucination in audio-visual LLMs. Our benchmark diagnoses speech-vision hallucinations from two critical and complementary aspects: semantic and temporal. Experimental results demonstrate that state-of-the-art open-source audio-visual LLMs struggle with aligning speech content with corresponding visual signals, with a near-random accuracy on multiple tasks. In contrast, Gemini 2.5 Pro significantly outperforms the open-source models. Our analysis suggests that their failures stem from limited ability in cross-modality understanding, despite strong performance in single-modality perception. Our work uncovers a new and fundamental limitation of current audio-visual LLMs and highlights the need for speech-grounded video comprehension. Project page: https://chenshuang-zhang.github.io/projects/svhalluc/. --- ### CARVE: Certified Affordable Repair of Vetoed Maneuvers via Envelopes for Interactive Driving Authors: Yifan Wang Published: 2026-05-31 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2606.02641v1 Interactive driving exposes a failure mode that is easy to miss in rule-aware autonomous-driving stacks: a hard-rule margin can be negative for an ego candidate even though a small lawful accommodation by a non-priority agent would restore feasibility. Existing rulebooks, shields, and reachability filters are strong at vetoing unsafe actions, while prediction-based planners model likely responses. Neither returns a runtime proof object that states which bounded multi-agent edit repairs the maneuver, who owns the edit, whether the request is right-of-way affordable, and what ego fallback remains if the request is not observed. We formulate this missing object as *interactive repair certification* and introduce *CARVE*, a prediction-free certificate layer over a finite lattice of ego-owned and agent-owned tactical operators. Agent-owned requests are admissible only inside \(B_j(s) = β(π_j)α_j^{\max}(s)\), a cooperation envelope that separates kinematic reachability from normative priority. The resulting certificate records the binding rule, repair category, repair set, responsibility-weighted cost split, and fallback. On 589 Lanelet2-geometry-grounded INTERACTION replay episodes, CARVE-Greedy accepts 98.64% of initially vetoed maneuvers and recovers 370/378 human-resolved false vetoes, while preserving 589/589 right-of-way respect, zero priority-agent false positives, and 400/400 negative-stress vetoes. We prove certificate soundness, structural right-of-way respect, exact finite-lattice minimality, fallback contingency, and blame-consistency conditions. CARVE does not predict or require another driver's compliance; it certifies whether a proposed interaction is bounded, attributable, and normatively admissible under declared assumptions. --- ### D-Judge: Disrupting Multi-Turn Jailbreaks using Semantics-Preserving Output Rewriting Authors: Huanli Gong, Zhipeng Wei, Yu Fu, Haz Sameen Shahgir, Ananya Gupta, Yue Dong, N. Benjamin Erichson Published: 2026-05-31 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2606.02640v1 Multi-turn jailbreak attacks pose a growing threat to large language model (LLM) safety because they exploit feedback from auxiliary judge models to iteratively refine prompts toward harmful goals. Existing defenses largely detect or block unsafe content at individual turns or at the final response, leaving the judge-driven refinement loop intact and allowing attackers to extract informative feedback from intermediate interactions. We introduce D-Judge, a semantics-preserving output rewriting defense that intervenes directly in this loop by rewriting the victim LLM's responses before they are evaluated by the attacker's judge. By misaligning the judge's feedback signal without changing the meaning of the original response, D-Judge derails the attacker's prompt-refinement process, causing subsequent queries to be optimized against a distorted signal of attack progress. To improve D-Judge's ability to produce such rewrites, we construct a dataset of semantically equivalent response pairs that induce different judge-assigned harmfulness scores, and use it for supervised fine-tuning followed by direct preference optimization. Experiments on HarmBench show that D-Judge reduces the success rate of state-of-the-art multi-turn jailbreaks while preserving performance on benign benchmarks. --- ### Sparse-View Lung Nodule Volumetry from Digitally Reconstructed Radiographs via AReT: Anatomy-Regularized TensoRF Authors: Spoorthi M, Suja Palaniswamy Published: 2026-05-31 Categories: eess.IV, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2606.02639v1 We identify and resolve a previously unreported failure mode in TensoRF when applied to X-ray attenuation fields: the default density shift of -10, originally introduced for RGB scene reconstruction, suppresses density gradients and prevents sparse-view medical reconstruction regardless of learning rate or regularization strategy. Setting the density shift to zero restores gradient flow and enables stable volumetric reconstruction of pulmonary nodules from only three orthogonal X-ray projections. Building on this, we propose AReT, an anatomy-regularized tensorial radiance field framework for lung nodule reconstruction using coronal, sagittal, and axial projections from the LIDC-IDRI dataset (19 patients, radiologist-annotated nodules). Unlike existing NeRF approaches requiring dense multi-view acquisition, AReT is designed for sparse-view thoracic imaging and incorporates chest-anatomy-aware regularization combining L1 sparsity and total variation smoothness. A systematic comparison across 11 reconstruction strategies shows anatomy-aware regularization consistently outperforms generative-prior-guided approaches. Evaluated against radiologist consensus segmentations, AReT achieves Pearson r=0.983 (p<0.0001) for clinically actionable nodules >=10 mm (n=14), median absolute volumetric error of 11.4%, near-zero systematic bias of -77.3 mm^3, and 8.4x improvement over spherical volume approximation. --- ### SegTune: Structured and Fine-Grained Control for Song Generation Authors: Yuejiao Wang, Zihao Ji, Pengfei Cai, Xu Li, Haorui Zheng, Zewen Song, Zhongliang Liu, Chen Zhang, Pengfei Wan Published: 2026-05-31 Categories: cs.SD, cs.AI, eess.AS arXiv: https://arxiv.org/abs/2606.02638v1 Recent advances in neural song generation have enabled high-quality synthesis from lyrics and global textual prompts. However, most systems fail to model temporally varying attributes of songs, severely limiting fine-grained control over musical structure and dynamics. To address this, we propose SegTune, a Diffusion Transformer-based framework enabling structured and fine-grained controllability by allowing users or large language models (LLMs) to specify local musical descriptions aligned to song segments. These segment prompts are temporally broadcast to corresponding time windows, while global prompts ensure stylistic coherence. To support precise lyric-to-music alignment, we introduce an LLM-based duration predictor that autoregressively generates sentence-level timestamps in LyRiCs format. We further construct a large-scale data pipeline for high-quality song collection with aligned lyrics and prompts, and propose new metrics to evaluate segment alignment and vocal consistency. Experiments demonstrate that SegTune outperforms existing baselines in both musicality and controllability. Visit our project page (https://github.com/KlingAIResearch/SegTune) for codes and more generated songs. --- ### Semantic Retrieval for Product Search in E-Commerce Authors: Nikhil Kothari, Saksham Samdani, Ritam Mallick, Praveen Gupta, Ankit Vijay, Surender Kumar Published: 2026-05-31 Categories: cs.IR, cs.LG arXiv: https://arxiv.org/abs/2606.01504v1 Semantic retrieval in e-commerce must handle short, noisy, and colloquial queries over large product catalogs with fine-grained attribute distinctions. We present a Siamese LLM dual-encoder trained through a two-stage pipeline: contrastive learning with a false-negative margin mask to prevent penalization of near-duplicate products, followed by Relative Odds Alignment for Retrieval (ROAR), a preference optimization objective that extends Bradley-Terry to variable-sized graded relevance groups via consecutive odds-ratio margins. The training corpus mirrors this progression - substitute query-product pairs provide coarse semantic supervision in Stage 1 and graded relevance annotations drive fine-grained ranking in Stage 2. The resulting system accurately retrieves exact matches while correctly ordering substitutes and complementary products, with gains confirmed across query-frequency strata and business verticals, and statistical significance validated through live A/B deployment at scale. --- ### On the Limits of Token Reduction for Efficient Unified Vision Language Training Authors: Siyi Chen, Weiming Zhuang, Jingtao Li, Lingjuan Lv Published: 2026-05-31 Categories: cs.CV, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2606.01503v1 Unified vision-language models (VLMs) integrate visual understanding and visual generation within a single autoregressive backbone, but their joint training is computationally expensive and largely overlooked from an efficiency perspective. In this work, we study the feasibility and limits of token-reduction-based acceleration for unified VLM training. Through a systematic analysis of layerwise attention allocation, we uncover a fundamental asymmetry: visual understanding exhibits substantial late-layer visual redundancy, whereas visual generation maintains persistent dependence on image tokens across depth. Guided by this observation, we design task-specific accelerators that selectively reduce image-token computation for each objective. While these methods achieve significant efficiency gains in isolated settings, we observe a consistent synergy loss under unified training -- task-specific token dropping necessitates divergent parameter pathways and eliminates the mutual performance gains typically observed in joint optimization. Our findings suggest that efficient unified modeling requires preserving shared cross-task structures, highlighting the need for synergy-aware acceleration strategies. Project page: https://chicychen.github.io/TokenReductionUnifiedVLM/. --- ### Move the Query, Not the Cache: Characterizing Cross-Instance Latent Attention Redistribution Across GPU Fabrics Authors: Bole Ma, Jan Eitzinger, Harald Köstler, Gerhard Wellein Published: 2026-05-31 Categories: cs.DC, cs.AI, cs.NI arXiv: https://arxiv.org/abs/2606.01502v1 Frontier LLMs increasingly decide what a query attends to with a sparse-attention indexer that picks a few KV-cache blocks per query: attention's unit is now a small, reusable chunk. Agentic workloads hammer it: many sub-agents query one large codebase, reusing the same blocks. When that corpus outgrows one GPU it is partitioned across instances, so a query and the blocks it selects often sit on different GPUs: answering it means attention across instances. The reflex of prior cross-instance KV systems is to move the cache: pull the selected blocks to the requester. Multi-head Latent Attention inverts the arithmetic, compressing each token's key and value into one narrow vector, so a routed query row is only ~1 KB, smaller than the chunk it attends; routing the query is then often cheaper than moving the cache. Which primitive wins, over which fabric and request shape, is uncharted, least of all on device-initiated RDMA that makes per-request cross-node transfers cheap. We characterize cross-instance MLA attention on a real multi-node H100 cluster, distilling two reusable artifacts: a topology-aware cost model (probe / transfer / compute / return / merge) and a closed-form route/fetch/local predicate, whose constants we measure on real IBGDA, where the model tracks batched round-trips to within ~7%. At decode it routes the query, trading the cost of moving the cache (a ~3 ms re-adaptation splice for a contiguous chunk, or a scattered gather under selection) for a tens-of-microsecond round trip, and picks the fabric by probe latency, not peak bandwidth. We instantiate the cost model and predicate for MLA, but neither is MLA-specific: they apply wherever compression or sparse selection shrinks attention to small chunks (DeepSeek-V3.2, V4, and GLM-5.1 today). Extending them to a new architecture requires measuring just two coefficients: the routed payload and fetch's move-the-cache cost. --- ### TimeSage-MT: A Multi-Turn Benchmark for Evaluating Agentic Time Series Reasoning Authors: Yaxuan Kong, Qingren Yao, Yuqi Nie, Yichen Li, Yilei Shao, Stefan Zohren, Anna Vettoruzzo, Joaquin Vanschoren, Ming Jin, Qingsong Wen Published: 2026-05-31 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2606.01498v1 Time series data inform critical decisions across many real-world domains. While large language model (LLM) agents can analyze data through natural language and tools, it remains unclear whether they can conduct reliable time series analysis across multi-turn conversations. Existing benchmarks focus on single-step tasks such as forecasting and anomaly detection, overlooking practical workflows where user goals evolve, agents must build on prior analyses, and conclusions emerge from accumulated evidence. In this work, we introduce TimeSage-MT, a multi-turn benchmark for agentic time series reasoning with 240 tasks and 2,680 dialogue turns across 8 real-world domains, spanning basic exploration to decision-oriented analysis. TimeSage-MT is built through a reproducible pipeline that converts real-world time series data into multi-turn conversations with verifiable answers. It provides a unified evaluation protocol and public leaderboard for comparing time series agentic systems. To demonstrate the benchmark's utility, we evaluate frontier LLMs alongside TimeSage, a novel structured agent equipped with a comprehensive time series skill library. The results show sharp performance drops on decision-oriented tasks, driven by failures in memory, uncertainty handling, and domain-based decision making. TimeSage-MT exposes critical gaps in current agentic reasoning and provides a rigorous foundation for future development. --- ### CART: Context-Anchored Recurrent Transformer -- A Parameter-Efficient Architecture with Learned Stability Authors: Chad A. Capps Published: 2026-05-31 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2606.01495v2 We present CART (Context-Anchored Recurrent Transformer), a parameter-efficient language model that reuses a single shared core block R times across depth. Unlike prior looped transformers that recompute key-value tensors at every iteration, CART computes K and V once from a multi-layer prelude and has the recurrent core cross-attend to those frozen tensors via multi-head latent attention. A learned Linear Time-Invariant (LTI) gate keeps the recurrence stable: its spectral radius settles in a narrow band (rho in (0.79, 0.83)) across all 36 fully-trained configurations. We evaluate CART on single consumer GPUs in two stages: a 64-configuration screen at 3,000 steps, then 36 configurations (P=6, R in {6,8,10}, three seeds) trained for 30,500 steps (~1B tokens). Two patterns hold across widths d in {256,512,768,1024}: prelude depth P dominates loop count R, and the Stage-1 ranking of R reverses at full training (R=6 becomes best at d>=512). At the binding d=1024 parameter-parity test, CART does not beat a parameter-matched dense baseline, losing by 1-2% at stored-parameter parity and by ~10% at effective-parameter parity. Diagnostic ablations split the effective-parameter gap into ~5% from weight sharing and a residual ~5% from the heterogeneous prelude/anchor/core/coda framing; the recurrent-core machinery (hyper-connections, LTI gate, loop-index embedding) is individually vestigial. Variable-R inference degrades on both sides of the trained R, a negative result for test-time depth scaling under this recipe. --- ### ClawHub Security Signals: When VirusTotal, Static Analysis, and SkillSpector Disagree Authors: Vincent Koc, Patrick Erichsen, Jacob Tomlinson, Agustin Rivera, Michael Appel, Nir Paz Published: 2026-05-31 Categories: cs.CR, cs.AI, cs.SE arXiv: https://arxiv.org/abs/2606.01494v1 Agent skills extend AI agents with reusable instructions, tools, scripts, references, and workflows, establishing a security boundary distinct from both model safety and traditional package-malware detection. ClawHub Security Signals is a sanitized dataset of 67,453 latest public OpenClaw skill versions. Each row pairs redacted SKILL.md content and sanitized bundled files where present with a final ClawScan registry verdict and evidence from three scanner families: VirusTotal, static heuristic analysis, and NVIDIA SkillSpector. Rather than estimating malicious-skill prevalence, we study scanner disagreement. The three scanners rarely flag the same skills: any pair overlaps on at most 10.4% of their combined positives, only 0.69% of skills are flagged by all three, and 81.9% of flagged skills are identified by a single scanner. The disagreement is structured by attack surface. SkillSpector, which raises semantic agentic-risk advisories rather than malware-reputation signals, is positive for 19,209 of 25,504 suspicious rows (75.3%) but only 14 of 206 malicious rows (6.8%). The malicious-verdict region shows the inverse profile: 150 of 206 malicious rows (72.8%) are VirusTotal-positive, consistent with bundled-code malware evidence. These results show that agent-skill security requires layered governance, not single-scanner allow/block decisions. The corpus is released as a sanitized silver-standard dataset: labels are the registry's automated verdicts, not human-annotated ground truth, and the release represents an early, versioned snapshot intended to support the community while a human-annotated subset is developed. Further research is encouraged, including models tailored for skill-security triage. --- ### Splatshot: 3D Face Avatar Generation from a Single Unconstrained Photo Authors: Hao Liang, Zhixuan Ge, Soumendu Majee, Joanna Li, Ashok Veeraraghavan, Guha Balakrishnan Published: 2026-05-31 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.01493v1 Reconstructing a photorealistic 3D face avatar from a single unconstrained photograph is challenging: feed-forward 3D Gaussian Splatting (3DGS) models degrade on out-of-distribution inputs, while pretrained diffusion models produce high-fidelity images but lack multi-view consistency. We observe that these paradigms are fundamentally complementary: explicit 3D representations guarantee geometric consistency, whereas 2D diffusion priors ensure photorealism. Building on this, we propose SplatShot, a training-free framework that couples these representations directly within the denoising process. Given a base 3DGS face model and a single reference image, we jointly denoise all target views using a per-step 3D feedback loop. At each timestep, we predict clean images from the noisy latents, refit the 3DGS to these multi-view predictions, and back-propagate the photometric discrepancy between the 3DGS re-renderings and 2D predictions into the noise estimate. This steers the sampling trajectory toward strictly 3D-coherent, identity-faithful outputs. Experiments on diverse in-the-wild images demonstrate that SplatShot produces 3D avatars with superior identity preservation, photorealism, and multi-view consistency. --- ### LLM Consortium for Software Design Refinement: A Controlled Experiment on Multi-Agent Collaboration Topologies Authors: Nagarjuna Kanamarlapudi, Praveen K Published: 2026-05-31 Categories: cs.SE, cs.AI, cs.MA arXiv: https://arxiv.org/abs/2606.01490v1 We present a controlled experiment evaluating 12 multi-agent LLM collaboration topologies for software architecture design. Using a $2\times2\times2$ factorial design (Authority $\times$ Roles $\times$ Dynamics), we conducted 520 experimental runs across 8 design tasks of varying complexity, with 5 repetitions each. Designs were evaluated on a 12-dimensional rubric by three independent automated evaluators (GPT-OSS 120B, Claude Opus 4.6, Claude Sonnet 4.6). We report four core findings. First, structural adversarial (v4b) ranks #1 by ensemble -- a prompt-engineered adversarial variant that demands rewrite mandates rather than patches (weighted ensemble: 4.637/5.0). Second, cross-model review wins unanimously at #2 -- generate with one model, review with another -- ranking #2 by all three evaluators (weighted ensemble: 4.606). Third, evaluator diversity is itself a finding -- all three evaluators agree v4b is best and v3 is worst, but disagree sharply on v2b (Claude d=1.44 vs. GPT-OSS d=0.45), revealing how different model families weight design qualities. Fourth, parallel merge is fundamentally broken -- all three evaluators place merge variants in the bottom tier (3.65-3.79), due to token starvation and the Frankenstein effect. The weighted ensemble ($2\times$Opus + $2\times$Sonnet + $1\times$GPT-OSS) provides robust rankings across 520 runs, confirmed through independent cross-validation. --- ### Perception First: A Frontier Native-Video Model with Self-Consistency for Implicit Video Question Answering Authors: Ali Alavi Published: 2026-05-31 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2606.01485v1 We describe our submission to the VRR Challenge @ CVPR 2026, built on the \emph{ImplicitQA} / \emph{VRR-QA} benchmark~\cite{implicitqa}: multiple-choice video question answering in which answers are deliberately \emph{not} observable in any single frame and must be inferred from spatial layout, motion, depth, viewpoint, causality, and social context across discontinuous frames of creative video. We conduct a systematic, training-free study spanning open-source Video-LMMs (Qwen2.5-VL~\cite{qwen25vl}, Qwen3-VL~\cite{qwen3vl}, InternVL3, Gemma-3, and the RL-tuned video reasoners Video-R1~\cite{videor1} and VideoChat-R1.5~\cite{videochatr15}) and a battery of inference-time strategies (chain-of-thought, question decomposition, describe-then-reason cascades, audio transcripts, spatial state prompting, self-consistency~\cite{selfconsistency}, multi-model ensembling, and category routing). Our central finding is that this benchmark is \emph{perception-bound rather than reasoning-bound}: reasoning-side augmentations are neutral-to-harmful, whereas base-model perceptual capability and lightweight test-time denoising are the only reliable levers. A per-category error analysis localizes the difficulty to low-level perception -- relative depth, viewpoint, and counting are the hardest categories, while causal and social reasoning are nearly solved -- and a prompt that explicitly injects monocular depth cues to attack the weakest category \emph{lowers} test accuracy by $5.8$ points, confirming that the model needs a better \emph{percept}, not a better \emph{procedure}. ---