# Research Papers — 2026-04 · bundle page 108 of 233

> Full-text markdown of papers #5351–5400 of 11612 published in 2026-04, newest first. AI-readable mirrors of the full 11612-paper 2026-04 archive (50 papers per file).

Canonical page: https://aipod.dev/papers/archive/2026-04/page/108
[← Previous bundle](https://aipod.dev/papers/archive/2026-04/page/107.md)
[Next bundle →](https://aipod.dev/papers/archive/2026-04/page/109.md)

### Can LLMs Accurately Score Medical Diagnoses and Clinical Reasoning?

Authors: Amy Rouillard, Sitwala Mundia, Linda Camara, Ziyaad Dangor, Michael Cameron Gramanie, Ismail Kalla, Shabir A. Madhi, Kajal Morar, Marlvin T. Ncube, Haroon Saloojee, Bruce A. Bassett
Published: 2026-04-16
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2604.14892v3

Evaluating medical AI systems using expert clinician panels is costly and slow, motivating the use of large language models (LLMs) as alternative adjudicators. Here, we evaluate an LLM Jury, composed of three frontier AI models, for scoring 3334 diagnoses on 300 real-world low- and middle-income country (LMIC) hospital cases. Both LLM- and clinician-generated diagnoses are scored against expert panel diagnoses across four dimensions: diagnosis, differential diagnosis, clinical reasoning, and negative treatment risk. The LLM Jury scores are compared with expert and independent re-scoring panel scores to assess error metrics, inter-rater agreement, severe-risk errors, and the effect of post hoc calibration using isotonic regression. In our data, we find that: (i) the uncalibrated LLM Jury scores preserve ordinal agreement with the expert clinician panel scores, but are systematically lower; (ii) the probability of severe-risk errors is lower for the LLM Jury than the human expert re-score panels; (iii) the LLM Jury combined with LLM diagnoses can be used to identify diagnoses at high risk of error, enabling targeted expert review and improved panel efficiency; (iv) the calibrated LLM Jury scores and rankings of diagnosing agents show excellent agreement with those of the primary expert panels; (v) LLM Jury models show no self-preference bias, they did not score diagnoses generated by their own underlying model or models from the same vendor more (or less) favourably than those generated by other models. Together, these results provide evidence that a calibrated LLM Jury is a trustworthy and reliable proxy for expert clinician evaluation in medical AI benchmarking. Confirming these findings in other clinical settings is an important direction for future work.

---

### MemoSight: Unifying Context Compression and Multi Token Prediction for Reasoning Acceleration

Authors: Xinyu Liu, Xin Liu, Bo Jin, Runsong Zhao, Pengcheng Huang, Junhao Ruan, Bei Li, Chunyang Xiao, Chenglong Wang, Tong Xiao, Jingbo Zhu
Published: 2026-04-16
Categories: cs.AI
arXiv: https://arxiv.org/abs/2604.14889v2

While chain-of-thought (CoT) reasoning enables LLMs to solve challenging reasoning tasks, the linear growth of the KV cache leads to substantial memory and inference overhead. Existing approaches such as context compression and multi-token prediction (MTP) improve efficiency from two complementary directions by compressing historical tokens and generating future tokens in parallel. However, effectively combining them remains challenging due to their different training paradigms and architectural assumptions. In this work, we propose MemoSight (Memory-Foresight-Based Reasoning), a unified framework that integrates context compression and MTP to improve inference efficiency while preserving CoT performance. MemoSight adopts a shared minimalist design based on special tokens and token-specific positional layouts for both compression and parallel prediction. Experiments on four reasoning benchmarks show that, compared to the vanilla SFT baseline, MemoSight reduces KV cache usage by up to 66% and improves inference speed by 56%, while incurring less than a 3% drop in average reasoning accuracy, yielding a better efficiency-accuracy trade-off than existing CoT compression methods.

---

### Reasoning Dynamics and the Limits of Monitoring Modality Reliance in Vision-Language Models

Authors: Danae Sánchez Villegas, Samuel Lewis-Lim, Nikolaos Aletras, Desmond Elliott
Published: 2026-04-16
Categories: cs.CL, cs.AI, cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2604.14888v3

Recent advances in vision language models (VLMs) offer reasoning capabilities, yet how these unfold and integrate visual and textual information remains unclear. We analyze reasoning dynamics in 18 VLMs covering instruction-tuned and reasoning-trained models from two different model families. We track confidence over Chain-of-Thought (CoT), measure the corrective effect of reasoning, and evaluate the contribution of intermediate reasoning steps. We find that models are prone to answer inertia, in which early commitments to a prediction are reinforced, rather than revised during reasoning steps. While reasoning-trained models show stronger corrective behavior, their gains depend on modality conditions, from text-dominant to vision-only settings. Using controlled interventions with misleading textual cues, we show that models are consistently influenced by these cues even when visual evidence is sufficient, and assess whether this influence is recoverable from CoT. Although this influence can appear in the CoT, its detectability varies across models and depends on what is being monitored. Reasoning-trained models are more likely to explicitly refer to the cues, but their longer and fluent CoTs can still appear visually grounded while actually following textual cues, obscuring modality reliance. In contrast, instruction-tuned models refer to the cues less explicitly, but their shorter traces reveal inconsistencies with the visual input. Taken together, these findings indicate that CoT provides only a partial view of how different modalities drive VLM decisions, with important implications for the transparency and safety of multimodal systems.

---

### Cooperate to Compete: Strategic Data Generation and Incentivization Framework for Coopetitive Cross-Silo Federated Learning

Authors: Thanh Linh Nguyen, Nguyen Van Huynh, Quoc-Viet Pham
Published: 2026-04-16
Categories: cs.AI, cs.DC, cs.GT
arXiv: https://arxiv.org/abs/2604.14886v1

In data-sensitive domains such as healthcare, cross-silo federated learning (CFL) allows organizations to collaboratively train AI models without sharing raw data. However, practical CFL deployments are inherently coopetitive, in which organizations cooperate during model training while competing in downstream markets. In such settings, training contributions, including data volume, quality, and diversity, can improve the global model yet inadvertently strengthen rivals. This dilemma is amplified by non-IID data, which leads to asymmetric learning gains and undermines sustained participation. While existing competition-aware CFL and incentive-design approaches reward organizations based on marginal training contributions, they fail to account for the costs of strengthening competitors. In this paper, we introduce CoCoGen+, a coopetition-compatible data generation and incentivization framework that jointly models non-IID data and inter-organizational competition while endogenizing GenAI-based synthetic data generation as a strategic decision. Specifically, CoCoGen+ formulates each training round as a weighted potential game, where organizations strategically decide how much synthetic data to generate by balancing learning performance gains against computational costs and competition-caused utility losses. We then provide a tractable equilibrium characterization and derive implementable generation strategies to maximize social welfare. To promote long-term collaboration, we integrate a payoff redistribution-based incentive mechanism to compensate organizations for their contributions and competition-caused utility degradation. Experiments on varying learning tasks validate the feasibility of CoCoGen+. The results show how non-IID data, competition intensity, and incentives shape organizational strategies and social welfare, while CoCoGen+ outperforms baselines in efficiency.

---

### RACER: Retrieval-Augmented Contextual Rapid Speculative Decoding

Authors: Zihong Zhang, Zuchao Li, Lefei Zhang, Ping Wang, Hai Zhao
Published: 2026-04-16
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2604.14885v1

Autoregressive decoding in Large Language Models (LLMs) generates one token per step, causing high inference latency. Speculative decoding (SD) mitigates this through a guess-and-verify strategy, but existing training-free variants face trade-offs: retrieval-based drafts break when no exact match exists, while logits-based drafts lack structural guidance. We propose $\textbf{RACER}$ ($\textbf{R}$etrieval-$\textbf{A}$ugmented $\textbf{C}$ont$\textbf{e}$xtual $\textbf{R}$apid Speculative Decoding), a lightweight and training-free method that integrates retrieved exact patterns with logit-driven future cues. This unification supplies both reliable anchors and flexible extrapolation, yielding richer speculative drafts. Experiments on Spec-Bench, HumanEval, and MGSM-ZH demonstrate that RACER consistently accelerates inference, achieving more than $2\times$ speedup over autoregressive decoding, and outperforms prior training-free methods, offering a scalable, plug-and-play solution for efficient LLM decoding. Our source code is available at $\href{https://github.com/hkr04/RACER}{https://github.com/hkr04/RACER}$.

---

### FSDETR: Frequency-Spatial Feature Enhancement for Small Object Detection

Authors: Jianchao Huang, Fengming Zhang, Haibo Zhu, Tao Yan
Published: 2026-04-16
Categories: cs.CV
arXiv: https://arxiv.org/abs/2604.14884v1

Small object detection remains a significant challenge due to feature degradation from downsampling, mutual occlusion in dense clusters, and complex background interference. To address these issues, this paper proposes FSDETR, a frequency-spatial feature enhancement framework built upon the RT-DETR baseline. By establishing a collaborative modeling mechanism, the method effectively leverages complementary structural information. Specifically, a Spatial Hierarchical Attention Block (SHAB) captures both local details and global dependencies to strengthen semantic representation. Furthermore, to mitigate occlusion in dense scenes, the Deformable Attention-based Intra-scale Feature Interaction (DA-AIFI) focuses on informative regions via dynamic sampling. Finally, the Frequency-Spatial Feature Pyramid Network (FSFPN) integrates frequency filtering with spatial edge extraction via the Cross-domain Frequency-Spatial Block (CFSB) to preserve fine-grained details. Experimental results show that with only 14.7M parameters, FSDETR achieves 13.9% APS on VisDrone 2019 and 48.95% AP50 tiny on TinyPerson, showing strong performance on small-object benchmarks. The code and models are available at https://github.com/YT3DVision/FSDETR.

---

### xFODE: An Explainable Fuzzy Additive ODE Framework for System Identification

Authors: Ertugrul Kececi, Tufan Kumbasar
Published: 2026-04-16
Categories: cs.LG
arXiv: https://arxiv.org/abs/2604.14883v1

Recent advances in Deep Learning (DL) have strengthened data-driven System Identification (SysID), with Neural and Fuzzy Ordinary Differential Equation (NODE/FODE) models achieving high accuracy in nonlinear dynamic modeling. Yet, system states in these frameworks are often reconstructed without clear physical meaning, and input contributions to the state derivatives remain difficult to interpret. To address these limitations, we propose Explainable FODE (xFODE), an interpretable SysID framework with integrated DL-based training. In xFODE, we define states in an incremental form to provide them with physical meanings. We employ fuzzy additive models to approximate the state derivative, thereby enhancing interpretability per input. To provide further interpretability, Partitioning Strategies (PSs) are developed, enabling the training of fuzzy additive models with explainability. By structuring the antecedent space during training so that only two consecutive rules are activated for any given input, PSs not only yield lower complexity for local inference but also enhance the interpretability of the antecedent space. To train xFODE, we present a DL framework with parameterized membership function learning that supports end-to-end optimization. Across benchmark SysID datasets, xFODE matches the accuracy of NODE, FODE, and NLARX models while providing interpretable insights.

---

### An Intelligent Robotic and Bio-Digestor Framework for Smart Waste Management

Authors: Radhika Khatri, Adit Tewari, Nikhil Sharma, M. B. Srinivas
Published: 2026-04-16
Categories: cs.RO, cs.LG
arXiv: https://arxiv.org/abs/2604.14882v1

Rapid urbanization and continuous population growth have made municipal solid waste management increasingly challenging. These challenges highlight the need for smarter and automated waste management solutions. This paper presents the design and evaluation of an integrated waste management framework that combines two connected systems, a robotic waste segregation module and an optimized bio-digestor. The robotic waste segregation system uses a MyCobot 280 Jetson Nano robotic arm along with YOLOv8 object detection and robot operating system (ROS)-based path planning to identify and sort waste in real time. It classifies waste into four different categories with high precision, reducing the need for manual intervention. After segregation, the biodegradable waste is transferred to a bio-digestor system equipped with multiple sensors. These sensors continuously monitor key parameters, including temperature, pH, pressure, and motor revolutions per minute. The Particle Swarm Optimization (PSO) algorithm, combined with a regression model, is used to dynamically adjust system parameters. This intelligent optimization approach ensures stable operation and maximizes digestion efficiency under varying environmental conditions. System testing under dynamic conditions demonstrates a sorting accuracy of 98% along with highly efficient biological conversion. The proposed framework offers a scalable, intelligent, and practical solution for modern waste management, making it suitable for both residential and industrial applications.

---

### The Missing Knowledge Layer in AI: A Framework for Stable Human-AI Reasoning

Authors: Rikard Rosenbacke, Carl Rosenbacke, Victor Rosenbacke, Martin McKee
Published: 2026-04-16
Categories: cs.AI, cs.CY, cs.HC
arXiv: https://arxiv.org/abs/2604.14881v1

Large language models are increasingly integrated into decision-making in areas such as healthcare, law, finance, engineering, and government. Yet they share a critical limitation: they produce fluent outputs even when their internal reasoning has drifted. A confident answer can conceal uncertainty, speculation, or inconsistency, and small changes in phrasing can lead to different conclusions. This makes LLMs useful assistants but unreliable partners in high-stakes contexts. Humans exhibit a similar weakness, often mistaking fluency for reliability. When a model responds smoothly, users tend to trust it, even when both model and user are drifting together. This paper is the first in a five-paper research series on stabilising human-AI reasoning. The series proposes a two-layer approach: Parts II-IV introduce human-side mechanisms such as uncertainty cues, conflict surfacing, and auditable reasoning traces, while Part V develops a model-side Epistemic Control Loop (ECL) that detects instability and modulates generation accordingly. Together, these layers form a missing operational substrate for governance by increasing signal-to-noise at the point of use. Stabilising interaction makes uncertainty and drift visible before enforcement is applied, enabling more precise capability governance. This aligns with emerging compliance expectations, including the EU AI Act and ISO/IEC 42001, by making reasoning processes traceable under real conditions of use. The central claim is that fluency is not reliability. Without structures that stabilise both human and model reasoning, AI cannot be trusted or governed where it matters most.

---

### xFODE+: Explainable Type-2 Fuzzy Additive ODEs for Uncertainty Quantification

Authors: Ertugrul Kececi, Tufan Kumbasar
Published: 2026-04-16
Categories: cs.LG, eess.SY
arXiv: https://arxiv.org/abs/2604.14880v1

Recent advances in Deep Learning (DL) have boosted data-driven System Identification (SysID), but reliable use requires Uncertainty Quantification (UQ) alongside accurate predictions. Although UQ-capable models such as Fuzzy ODE (FODE) can produce Prediction Intervals (PIs), they offer limited interpretability. We introduce Explainable Type-2 Fuzzy Additive ODEs for UQ (xFODE+), an interpretable SysID model which produces PIs alongside point predictions while retaining physically meaningful incremental states. xFODE+ implements each fuzzy additive model with Interval Type-2 Fuzzy Logic Systems (IT2-FLSs) and constraints membership functions to the activation of two neighboring rules, limiting overlap and keeping inference locally transparent. The type-reduced sets produced by the IT2-FLSs are aggregated to construct the state update together with the PIs. The model is trained in a DL framework via a composite loss that jointly optimizes prediction accuracy and PI quality. Results on benchmark SysID datasets show that xFODE+ matches FODE in PI quality and achieves comparable accuracy, while providing interpretability.

---

### SOLIS: Physics-Informed Learning of Interpretable Neural Surrogates for Nonlinear Systems

Authors: Murat Furkan Mansur, Tufan Kumbasar
Published: 2026-04-16
Categories: cs.LG, cs.AI, eess.SY
arXiv: https://arxiv.org/abs/2604.14879v1

Nonlinear system identification must balance physical interpretability with model flexibility. Classical methods yield structured, control-relevant models but rely on rigid parametric forms that often miss complex nonlinearities, whereas Neural ODEs are expressive yet largely black-box. Physics-Informed Neural Networks (PINNs) sit between these extremes, but inverse PINNs typically assume a known governing equation with fixed coefficients, leading to identifiability failures when the true dynamics are unknown or state-dependent. We propose \textbf{SOLIS}, which models unknown dynamics via a \emph{state-conditioned second-order surrogate model} and recasts identification as learning a Quasi-Linear Parameter-Varying (Quasi-LPV) representation, recovering interpretable natural frequency, damping, and gain without presupposing a global equation. SOLIS decouples trajectory reconstruction from parameter estimation and stabilizes training with a cyclic curriculum and \textbf{Local Physics Hints} windowed ridge-regression anchors that mitigate optimization collapse. Experiments on benchmarks show accurate parameter-manifold recovery and coherent physical rollouts from sparse data, including regimes where standard inverse methods fail.

---

### GenRec: A Preference-Oriented Generative Framework for Large-Scale Recommendation

Authors: Yanyan Zou, Junbo Qi, Lunsong Huang, Yu Li, Kewei Xu, Jiabao Gao, Binglei Zhao, Xuanhua Yang, Sulong Xu, Shengjie Li
Published: 2026-04-16
Categories: cs.IR, cs.AI
arXiv: https://arxiv.org/abs/2604.14878v1

Generative Retrieval (GR) offers a promising paradigm for recommendation through next-token prediction (NTP). However, scaling it to large-scale industrial systems introduces three challenges: (i) within a single request, the identical model inputs may produce inconsistent outputs due to the pagination request mechanism; (ii) the prohibitive cost of encoding long user behavior sequences with multi-token item representations based on semantic IDs, and (iii) aligning the generative policy with nuanced user preference signals. We present GenRec, a preference-oriented generative framework deployed on the JD App that addresses above challenges within a single decoder-only architecture. For training objective, we propose Page-wise NTP task, which supervises over an entire interaction page rather than each interacted item individually, providing denser gradient signal and resolving the one-to-many ambiguity of point-wise training. On the prefilling side, an asymmetric linear Token Merger compresses multi-token Semantic IDs in the prompt while preserving full-resolution decoding, reducing input length by ~2X with negligible accuracy loss. To further align outputs with user satisfaction, we introduce GRPO-SR, a reinforcement learning method that pairs Group Relative Policy Optimization with NLL regularization for training stability, and employs Hybrid Rewards combining a dense reward model with a relevance gate to mitigate reward hacking. In month-long online A/B tests serving production traffic, GenRec achieves 9.5% improvement in click count and 8.7% in transaction count over the existing pipeline.

---

### Does RL Expand the Capability Boundary of LLM Agents? A PASS@(k,T) Analysis

Authors: Zhiyuan Zhai, Wenjing Yan, Xiaodan Shao, Xin Wang
Published: 2026-04-16
Categories: cs.LG
arXiv: https://arxiv.org/abs/2604.14877v1

Does reinforcement learning genuinely expand what LLM agents can do, or merely make them more reliable? For static reasoning, recent work answers the second: base and RL pass@k curves converge at large k. We ask whether this holds for agentic tool use, where T rounds of interaction enable compositional strategies that re-sampling cannot recover. We introduce PASS@(k,T), a two-dimensional metric that jointly varies sampling budget k and interaction depth T, separating capability expansion from efficiency improvement. Our main finding is that, contrary to the static-reasoning result, tool-use RL genuinely enlarges the capability boundary: the RL agent's pass-curve pulls above the base model's and the gap widens at large k rather than converging. The expansion is specific to compositional, sequential information gathering; on simpler tasks RL behaves as prior work predicts. Under matched training data, supervised fine-tuning regresses the boundary on the same compositional tasks, isolating self-directed exploration as the causal factor. Mechanism analysis shows RL reweights the base strategy distribution toward the subset whose downstream reasoning more often yields a correct answer, with the improvement concentrated on how the agent integrates retrieved information. These results reconcile optimistic and pessimistic readings of RL for LLMs: both are correct, on different task types.

---

### Regret Tail Characterization of Optimal Bandit Algorithms with Generic Rewards

Authors: Subhodip Panda, Shubhada Agrawal
Published: 2026-04-16
Categories: cs.IT, cs.LG
arXiv: https://arxiv.org/abs/2604.14876v1

We study the tail behavior of regret in stochastic multi-armed bandits for algorithms that are asymptotically optimal in expectation. While minimizing expected regret is the classical objective, recent work shows that even such algorithms can exhibit heavy regret tails, incurring large regret with non-negligible probability. Existing sharp characterizations of regret tails are largely restricted to parametric settings, such as single-parameter exponential families. In this work, we extend the $\KLinf$-UCB algorithm of to a broad nonparametric class of reward distributions satisfying mild assumptions, and establish its asymptotic optimality in expectation. We then analyze the tail behavior of its regret and derive a novel upper bound on the regret tail probability. As special cases, our results recover regret-tail guarantees for both bounded-support and heavy-tailed (moment-bounded) bandit models. Moreover, for the special case of finitely-supported reward distributions, our upper bound matches the known lower bound exactly. Our results thus provide a unified and tight characterization of regret tails for asymptotically optimal KL-based UCB algorithms, going beyond parametric models.

---

### Open-Set Vein Biometric Recognition with Deep Metric Learning

Authors: Paweł Pilarek, Marcel Musiałek, Anna Górska
Published: 2026-04-16
Categories: cs.CV
arXiv: https://arxiv.org/abs/2604.14874v2

Most state-of-the-art vein recognition methods rely on closed-set classification, which inherently limits their scalability and prevents the adaptive enrollment of new users without complete model retraining. We rigorously evaluate the computational boundaries of Deep Metric Learning (DML) under strict open-set constraints. Unlike standard closed-set approaches, we analyze the impact of data scarcity and domain shift on recognition performance. Our approach learns discriminative L2-normalised embeddings and employs prototype-based matching with a calibrated similarity threshold to effectively distinguish between enrolled users and unseen impostors. We evaluate the framework under a strict subject-disjoint protocol across four diverse datasets covering finger, wrist, and dorsal hand veins (MMCBNU 6000, UTFVP, FYO, and a dorsal hand-vein dataset). On the large-scale MMCBNU 6000 benchmark, our best model (ResNet50-CBAM) achieves an OSCR of 0.9945, AUROC of 0.9974, and EER of 1.57%, maintaining high identification accuracy (99.6% Rank-1) while robustly rejecting unknown subjects. Cross-dataset experiments evaluate the framework's generalisation across different acquisition setups, confirming that while the model handles large-scale data robustly, performance remains sensitive to domain shifts in low-data regimes. Ablation studies demonstrate that triplet-based objectives combined with a simple 1-NN classifier offer an optimal trade-off between accuracy and efficiency, enabling real-time deployment on commodity hardware.

---

### Curvature-Aligned Probing for Local Loss-Landscape Stabilization

Authors: Nikita Kiselev, Andrey Grabovoy
Published: 2026-04-16
Categories: cs.LG
arXiv: https://arxiv.org/abs/2604.14870v1

Local loss-landscape stabilization under sample growth is typically measured either pointwise or through isotropic averaging in the full parameter space. Despite practical value, both choices probe directions that contribute little to the dominant local deformation of strongly anisotropic neural landscapes. We recast stabilization as an observational problem and introduce a unified family of criteria parameterized by an aggregation order and a probing distribution; within this family we propose a curvature-aligned criterion $Δ_2^{(D)}$ that probes the loss increment field in the top-$D$ eigenspace of the empirical Hessian near a trained solution. Solely from a local quadratic model, we prove that $Δ_2^{(D)}$ preserves the $O(k^{-2})$ mean-squared rate of the full-space criterion while replacing ambient-dimension curvature dependence with dependence on the subspace dimension $D$; a corollary gives a closed-form spectral expression and a proposition identifies the top-$D$ eigenspace as extremal within the eigenspace-aligned family. We also derive scalable estimators based on Hessian-vector products, subspace Monte Carlo, and a closed-form Gaussian-moment proxy. On a decoder-only transformer, a curvature-aligned probe occupying a tiny fraction of parameter space already reproduces the full-space mean-squared signal to within numerical noise throughout the validated local regime, and the closed-form estimator is orders of magnitude faster than direct Monte Carlo after subspace construction.

---

### Vibe-Coding: Feedback-Based Automated Verification with no Human Code Inspection, a Feasibility Study

Authors: Michal Töpfer, František Plášil, Tomáš Bureš, Petr Hnětynka
Published: 2026-04-16
Categories: cs.SE, cs.AI
arXiv: https://arxiv.org/abs/2604.14867v1

Vibe coding inherently assumes iterative refinement of LLM-generated code through feedback loops. While effective for conventional software tasks, its reliability in runtime-adaptive systems is unclear -- especially when generated code is not manually inspected. This paper studies feedback-based automated verification of LLM-generated adaptation managers in Collective Adaptive Systems (CAS). We focus on the key challenges of verification in the loop: how to detect failures of generated code at runtime and how to report them precisely enough for an LLM to fix them. We combine the adaptation loop with a vibe-coding feedback loop where correctness is checked against (i) generic architectural constraints and (ii) functional constraints formalized in Functional Constraints Logic (FCL), a novel first-order temporal logic over potentially finite traces. Conducting the Dragon Hunt CAS case study, we show that fine-grained constraint violations provide actionable feedback that typically yields a valid adaptation manager within a few iterations, while simple coarse metric-based feedback often stalls. Our findings suggest that feedback precision is the dominant factor for reliable vibe coding in systems designed by domain experts with no programming skills, thereby obviating the need for human code inspection.

---

### MetaDent: Labeling Clinical Images for Vision-Language Models in Dentistry

Authors: Meng-Xun Li, Wen-Hui Deng, Zhi-Xing Wu, Chun-Xiao Jin, Jia-Min Wu, Yue Han, James Kit Hon Tsoi, Gui-Song Xia, Cui Huang
Published: 2026-04-16
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2604.14866v1

Vision-Language Models (VLMs) have demonstrated significant potential in medical image analysis, yet their application in intraoral photography remains largely underexplored due to the lack of fine-grained, annotated datasets and comprehensive benchmarks. To address this, we present MetaDent, a comprehensive resource that includes (1) a novel and large-scale dentistry image dataset collected from clinical, public, and web sources; (2) a semi-structured annotation framework designed to capture the hierarchical and clinically nuanced nature of dental photography; and (3) comprehensive benchmark suites for evaluating state-of-the-art VLMs on clinical image understanding. Our labeling approach combines a high-level image summary with point-by-point, free-text descriptions of abnormalities. This method enables rich, scalable, and task-agnostic representations. We curated 60,669 dental images from diverse sources and annotated a representative subset of 2,588 images using this meta-labeling scheme. Leveraging Large Language Models (LLMs), we derive standardized benchmarks: approximately 15K Visual Question Answering (VQA) pairs and an 18-class multi-label classification dataset, which we validated with human review and error analysis to justify that the LLM-driven transition reliably preserves fidelity and semantic accuracy. We then evaluate state-of-the-art VLMs across VQA, classification, and image captioning tasks. Quantitative results reveal that even the most advanced models struggle with a fine-grained understanding of intraoral scenes, achieving moderate accuracy and producing inconsistent or incomplete descriptions in image captioning. We publicly release our dataset, annotations, and tools to foster reproducible research and accelerate the development of vision-language systems for dental applications.

---

### Segment-Level Coherence for Robust Harmful Intent Probing in LLMs

Authors: Xuanli He, Bilgehan Sel, Faizan Ali, Jenny Bao, Hoagy Cunningham, Jerry Wei
Published: 2026-04-16
Categories: cs.CL, cs.CR
arXiv: https://arxiv.org/abs/2604.14865v1

Large Language Models (LLMs) are increasingly exposed to adaptive jailbreaking, particularly in high-stakes Chemical, Biological, Radiological, and Nuclear (CBRN) domains. Although streaming probes enable real-time monitoring, they still make systematic errors. We identify a core issue: existing methods often rely on a few high-scoring tokens, leading to false alarms when sensitive CBRN terms appear in benign contexts. To address this, we introduce a streaming probing objective that requires multiple evidence tokens to consistently support a prediction, rather than relying on isolated spikes. This encourages more robust detection based on aggregated signals instead of single-token cues. At a fixed 1% false-positive rate, our method improves the true-positive rate by 35.55% relative to strong streaming baselines. We further observe substantial gains in AUROC, even when starting from near-saturated baseline performance (AUROC = 97.40%). We also show that probing Attention or MLP activations consistently outperforms residual-stream features. Finally, even when adversarial fine-tuning enables novel character-level ciphers, harmful intent remains detectable: probes developed for the base LLMs can be applied ``plug-and-play'' to these obfuscated attacks, achieving an AUROC of over 98.85%.

---

### Schema-Key Wording as an Instruction Channel in Structured Generation under Constrained Decoding

Authors: Yifan Le
Published: 2026-04-16
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2604.14862v3

Constrained decoding is widely used to make large language models produce structured outputs that satisfy schemas such as JSON. Existing work mainly treats schemas as structural constraints, overlooking that schema-key tokens also enter the autoregressive context and may guide generation. To the best of our knowledge, we present the first systematic study of schema keys as an implicit instruction channel under constrained decoding. We formulate structured generation as a multi-channel instruction problem, where task signals can be placed in prompts, schema keys, or both. We further provide a projection-aware analysis that gives a sufficient condition under which an unconstrained expected-score advantage of an instructional key is preserved after grammar projection. Experiments on GSM8K and Math500 across seven language models show that changing only schema-key wording can substantially affect accuracy, with both positive and negative effects across models. Prompt-level and schema-level instructions also interact non-additively. The evidence is substantially stronger on GSM8K than on Math500. Our findings show that schema design is not merely output formatting, but part of instruction specification in structured generation.

---

### Schema Key Wording as an Instruction Channel in Structured Generation under Constrained Decoding

Authors: Yifan Le
Published: 2026-04-16
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2604.14862v2

Constrained decoding is widely used to make large language models produce structured outputs that satisfy schemas such as JSON. Existing work mainly treats schemas as structural constraints, overlooking that schema-key tokens also enter the autoregressive context and may guide generation. To the best of our knowledge, we present the first systematic study of schema keys as an implicit instruction channel under constrained decoding. We formulate structured generation as a multi-channel instruction problem, where task signals can be placed in prompts, schema keys, or both. We further provide a projection-aware analysis: a CoT-style key helps only when its semantic gain exceeds the distortion induced by grammar-constrained projection, offering a theoretical explanation for model-dependent key effects. Experiments on mathematical reasoning benchmarks show that changing only schema-key wording can substantially affect accuracy while keeping the prompt, model, output structure, and decoding setup fixed. Qwen models tend to benefit more from schema-level instructions, whereas LLaMA models rely more on prompt-level guidance, and the two channels interact non-additively. Our findings show that schema design is not merely output formatting, but part of instruction specification in structured generation.

---

### Best of both worlds: Stochastic & adversarial best-arm identification

Authors: Yasin Abbasi-Yadkori, Peter L. Bartlett, Victor Gabillon, Alan Malek, Michal Valko
Published: 2026-04-16
Categories: stat.ML, cs.LG
arXiv: https://arxiv.org/abs/2604.14860v1

We study bandit best-arm identification with arbitrary and potentially adversarial rewards. A simple random uniform learner obtains the optimal rate of error in the adversarial scenario. However, this type of strategy is suboptimal when the rewards are sampled stochastically. Therefore, we ask: Can we design a learner that performs optimally in both the stochastic and adversarial problems while not being aware of the nature of the rewards? First, we show that designing such a learner is impossible in general. In particular, to be robust to adversarial rewards, we can only guarantee optimal rates of error on a subset of the stochastic problems. We give a lower bound that characterizes the optimal rate in stochastic problems if the strategy is constrained to be robust to adversarial rewards. Finally, we design a simple parameter-free algorithm and show that its probability of error matches (up to log factors) the lower bound in stochastic problems, and it is also robust to adversarial ones.

---

### Benchmarks for Trajectory Safety Evaluation and Diagnosis in OpenClaw and Codex: ATBench-Claw and ATBench-Codex

Authors: Zhonghao Yang, Yu Li, Yanxu Zhu, Tianyi Zhou, Yuejin Xie, Haoyu Luo, Jing Shao, Xia Hu, Dongrui Liu
Published: 2026-04-16
Categories: cs.AI, cs.SE
arXiv: https://arxiv.org/abs/2604.14858v2

As agent systems move into increasingly diverse execution settings, trajectory-level safety evaluation and diagnosis require benchmarks that evolve with them. ATBench is a diverse and realistic agent trajectory benchmark for safety evaluation and diagnosis. This report presents ATBench-Claw and ATBench-Codex, two domain-customized extensions that carry ATBench into the OpenClaw and OpenAI Codex / Codex-runtime settings. The key adaptation mechanism is to analyze each new setting, customize the three-dimensional Safety Taxonomy over risk source, failure mode, and real-world harm, and then use that customized taxonomy to define the benchmark specification consumed by the shared ATBench construction pipeline. This extensibility matters because agent frameworks remain relatively stable at the architectural level even as their concrete execution settings, tool ecosystems, and product capabilities evolve quickly. Concretely, ATBench-Claw targets OpenClaw-sensitive execution chains over tools, skills, sessions, and external actions, while ATBench-Codex targets trajectories in the OpenAI Codex / Codex-runtime setting over repositories, shells, patches, dependencies, approvals, and runtime policy boundaries. Our emphasis therefore falls on taxonomy customization, domain-specific risk coverage, and benchmark design under a shared ATBench generation framework.

---

### ClimateCause: Complex and Implicit Causal Structures in Climate Reports

Authors: Liesbeth Allein, Nataly Pineda-Castañeda, Andrea Rocci, Marie-Francine Moens
Published: 2026-04-16
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2604.14856v1

Understanding climate change requires reasoning over complex causal networks. Yet, existing causal discovery datasets predominantly capture explicit, direct causal relations. We introduce ClimateCause, a manually expert-annotated dataset of higher-order causal structures from science-for-policy climate reports, including implicit and nested causality. Cause-effect expressions are normalized and disentangled into individual causal relations to facilitate graph construction, with unique annotations for cause-effect correlation, relation type, and spatiotemporal context. We further demonstrate ClimateCause's value for quantifying readability based on the semantic complexity of causal graphs underlying a statement. Finally, large language model benchmarking on correlation inference and causal chain reasoning highlights the latter as a key challenge.

---

### Adaptive Test-Time Compute Allocation for Reasoning LLMs via Constrained Policy Optimization

Authors: Zhiyuan Zhai, Bingcong Li, Bingnan Xiao, Ming Li, Xin Wang
Published: 2026-04-16
Categories: cs.LG
arXiv: https://arxiv.org/abs/2604.14853v1

Test-time compute scaling, the practice of spending extra computation during inference via repeated sampling, search, or extended reasoning, has become a powerful lever for improving large language model performance. Yet deploying these techniques under finite inference budgets requires a decision that current systems largely ignore: which inputs deserve more compute, and which can be answered cheaply? We formalize this as a constrained optimization problem (maximize expected accuracy subject to an average compute budget) and solve it with a two-stage Solve-then-Learn pipeline. In the solve stage, Lagrangian relaxation decomposes the global constraint into per-instance sub-problems, each admitting a closed-form oracle action that optimally prices accuracy against cost. We prove that the induced cost is monotone in the dual variable, enabling exact budget targeting via binary search. In the learn stage, a lightweight classifier is trained to predict oracle actions from cheap input features, amortizing the allocation rule for real-time deployment. We establish that the task-level regret of the learned policy is bounded by its imitation error times the worst-case per-instance gap, yielding a clean reduction from constrained inference to supervised classification. Experiments on MATH and GSM8K with three LLMs (DeepSeek-V3, GPT-4o-mini, Qwen2.5-7B) show that our method consistently outperforms uniform and heuristic allocation baselines, achieving up to 12.8% relative accuracy improvement on MATH under matched budget constraints, while closely tracking the Lagrangian oracle upper bound with over 91% imitation accuracy.

---

### Efficient Search of Implantable Adaptive Cells for Medical Image Segmentation

Authors: Emil Benedykciuk, Marcin Denkowski, Grzegorz M. Wójcik
Published: 2026-04-16
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2604.14849v1

Purpose: Adaptive skip modules can improve medical image segmentation, but searching for them is computationally costly. Implantable Adaptive Cells (IACs) are compact NAS modules inserted into U-Net skip connections, reducing the search space compared with full-network NAS. However, the original IAC framework still requires a 200-epoch differentiable search for each backbone and dataset. Methods: We analyzed the temporal behavior of operations and edges within IAC cells during differentiable search on public medical image segmentation benchmarks. We found that operations selected in the final discrete cell typically emerge among the strongest candidates early in training, and their architecture parameters stabilize well before the final epoch. Based on this, we propose a Jensen--Shannon-divergence-based stability criterion that tracks per-edge operation-importance distributions and progressively prunes low-importance operations during search. The accelerated framework is called IAC-LTH. Results: Across four public benchmarks (ACDC, BraTS, KiTS, AMOS), several 2-D U-Net backbones, and a 2-D nnU-Net pipeline, IAC-LTH discovers IAC cells whose patient-level segmentation performance matches and sometimes slightly exceeds that of cells found by the original full-length search, while reducing wall-clock NAS cost by 3.7x to 16x across datasets and backbones. These results are consistent across architectures, benchmarks, and both non-augmented and augmented training settings, while preserving the gains of IAC-equipped U-Nets over strong attention-based and dense-skip baselines. Conclusion: Competitive IAC architectures can be identified from early-stabilizing operations without running the full search, making adaptive skip-module design more practical for medical image segmentation under realistic computational constraints.

---

### TrigReason: Trigger-Based Collaboration between Small and Large Reasoning Models

Authors: Yi Zhao, Yajuan Peng, Cam-Tu Nguyen, Zuchao Li, Xiaoliang Wang, Xiaoming Fu, Hai Zhao
Published: 2026-04-16
Categories: cs.AI
arXiv: https://arxiv.org/abs/2604.14847v1

Large Reasoning Models (LRMs) achieve strong performance on complex tasks through extended chains of thought but suffer from high inference latency due to autoregressive reasoning. Recent work explores using Small Reasoning Models (SRMs) to accelerate LRM inference. In this paper, we systematically characterize the capability boundaries of SRMs and identify three common types of reasoning risks: (1) path divergence, where SRMs lack the strategic ability to construct an initial plan, causing reasoning to deviate from the most probable path; (2) cognitive overload, where SRMs fail to solve particularly difficult steps; and (3) recovery inability, where SRMs lack robust self-reflection and error correction mechanisms. To address these challenges, we propose TrigReason, a trigger-based collaborative reasoning framework that replaces continuous polling with selective intervention. TrigReason delegates most reasoning to the SRM and activates LRM intervention only when necessary-during initial strategic planning (strategic priming trigger), upon detecting extraordinary overconfidence (cognitive offload trigger), or when reasoning falls into unproductive loops (intervention request trigger). The evaluation results on AIME24, AIME25, and GPQA-D indicate that TrigReason matches the accuracy of full LRMs and SpecReason, while offloading 1.70x - 4.79x more reasoning steps to SRMs. Under edge-cloud conditions, TrigReason reduces latency by 43.9\% and API cost by 73.3\%. Our code is available at \href{https://github.com/QQQ-yi/TrigReason}{https://github.com/QQQ-yi/TrigReason}

---

### Zero-Shot Retail Theft Detection via Orchestrated Vision Models: A Model-Agnostic, Cost-Effective Alternative to Trained Single-Model Systems

Authors: Haileab Yagersew
Published: 2026-04-16
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2604.14846v1

Retail theft costs the global economy over \$100 billion annually, yet existing AI-based detection systems require expensive custom model training on proprietary datasets and charge \$200-500/month per store. We present Paza, a zero-shot retail theft detection framework that achieves practical concealment detection without training any model. Our approach orchestrates multiple existing models in a layered pipeline - cheap object detection and pose estimation running continuously, with an expensive vision-language model (VLM) invoked only when behavioral pre-filters trigger. A multi-signal suspicion pre-filter (requiring dwell time plus at least one behavioral signal) reduces VLM invocations by 240x compared to per-frame analysis, bounding calls to <=10/minute and enabling a single GPU to serve 10-20 stores. The architecture is model-agnostic: the VLM component accepts any OpenAI-compatible endpoint, enabling operators to swap between models such as Gemma 4, Qwen3.5-Omni, GPT-4o, or future releases without code changes - ensuring the system improves as the VLM landscape evolves. We evaluate the VLM component on the DCSASS synthesized shoplifting dataset (169 clips, controlled environment), achieving 89.5% precision and 92.8% specificity at 59.3% recall zero-shot - where the recall gap is attributable to sparse frame sampling in offline evaluation rather than VLM reasoning failures, as precision and specificity are the operationally critical metrics determining false alarm rates. We present a detailed cost model showing viability at \$50-100/month per store (3-10x cheaper than commercial alternatives), and introduce a privacy-preserving design that obfuscates faces in the detection pipeline. The source code is available at https://github.com/xHaileab/Paza-AI.

---

### Exploring and Testing Skill-Based Behavioral Profile Annotation: Human Operability and LLM Feasibility under Schema-Guided Execution

Authors: Yufeng Wu
Published: 2026-04-16
Categories: cs.CL
arXiv: https://arxiv.org/abs/2604.14843v1

Behavioral Profile (BP) annotation is difficult to automate because it requires simultaneous coding across multiple linguistic dimensions. We treat BP annotation as a bundle of annotation skills rather than a single task and evaluate LLM-assisted BP annotation from this perspective. Using 3,134 concordance lines of 30 Chinese metaphorical color-term derivatives and a 14-feature BP schema, we implement a skill-file-driven pipeline in which each feature is externally defined through schema files, decision rules, and examples. Two human annotators completed a two-round schema-only protocol on a 300-instance validation subset, enabling BP skills to be classified as directly operable, recoverable under focused re-annotation, or structurally underspecified. GPT-5.4 and three locally deployable open-source models were then evaluated under the same setup. Results show that BP annotation is highly heterogeneous at the skill level: 5 skills are directly operable, 4 are recoverable after focused re-annotation, and 5 remain structurally underspecified. GPT-5.4 executes the retained skills with substantial reliability (accuracy = 0.678, \k{appa} = 0.665, weighted F1 = 0.695), but this feasibility is selective rather than global. Human and GPT difficulty profiles are strongly aligned at the skill level (r = 0.881), but not at the instance level (r = 0.016) or lexical-item level (r = -0.142), a pattern we describe as shared taxonomy, independent execution. Pairwise agreement further suggests that GPT is better understood as an independent third skill voice than as a direct human substitute. Open-source failures are concentrated in schema-to-skill execution problems. These findings suggest that automatic annotation should be evaluated in terms of skill feasibility rather than task-level automation.

---

### Intermediate Layers Encode Optimal Biological Representations in Single-Cell Foundation Models

Authors: Vincenzo Yuto Civale, Roberto Semeraro, Andrew David Bagdanov, Alberto Magi
Published: 2026-04-16
Categories: cs.AI
arXiv: https://arxiv.org/abs/2604.14838v1

Current single-cell foundation model benchmarks universally extract final layer embeddings, assuming these represent optimal feature spaces. We systematically evaluate layer-wise representations from scFoundation (100M parameters) and Tahoe-X1 (1.3B parameters) across trajectory inference and perturbation response prediction. Our analysis reveals that optimal layers are task-dependent (trajectory peaks at 60% depth, 31% above final layers) and context-dependent (perturbation optima shift 0-96% across T cell activation states). Notably, first-layer embeddings outperform all deeper layers in quiescent cells, challenging assumptions about hierarchical feature abstraction. These findings demonstrate that "where" to extract features matters as much as "what" the model learns, necessitating systematic layer evaluation tailored to biological task and cellular context rather than defaulting to final-layer embeddings.

---

### Improved Multiscale Structural Mapping with Supervertex Vision Transformer for the Detection of Alzheimer's Disease Neurodegeneration

Authors: Geonwoo Baek, David H. Salat, Ikbeom Jang
Published: 2026-04-16
Categories: cs.CV
arXiv: https://arxiv.org/abs/2604.14837v1

Alzheimer's disease (AD) confirmation often relies on positron emission tomography (PET) or cerebrospinal fluid (CSF) analysis, which are costly and invasive. Consequently, structural MRI biomarkers such as cortical thickness (CT) are widely used for non-invasive AD screening. Multiscale structural mapping (MSSM) was recently proposed to integrate gray-white matter contrasts (GWCs) with CT from a single T1-weighted MRI (T1w) scan. Building on this framework, we propose MSSM+, together with surface supervertex mapping (SSVM) and a Supervertex Vision Transformer (SV-ViT). 3D T1w images from individuals with AD and cognitively normal (CN) controls were analyzed. MSSM+ extends MSSM by incorporating sulcal depth and cortical curvature at the vertex level. SSVM partitions the cortical surface into supervertices (surface patches) that effectively represent inter- and intra-regional spatial relationships. SV-ViT is a Vision Transformer architecture operating on these supervertices, enabling anatomically informed learning from surface mesh representations. Compared with MSSM, MSSM+ identified more spatially extensive and statistically significant group differences between AD and CN. In AD vs. CN classification, MSSM+ achieved a 3%p higher area under the precision-recall curve than MSSM. Vendor-specific analyses further demonstrated reduced signal variability and consistently improved classification performance across MR manufacturers relative to CT, GWCs, and MSSM. These findings suggest that MSSM+ combined with SV-ViT is a promising MRI-based imaging marker for AD detection prior to CSF/PET confirmation.

---

### Beyond Literal Summarization: Redefining Hallucination for Medical SOAP Note Evaluation

Authors: Bhavik Vachhani, Kush Shrisvastava, Pranshu Nema, Sai Chiranthan
Published: 2026-04-16
Categories: cs.AI
arXiv: https://arxiv.org/abs/2604.14829v1

Evaluating large language models (LLMs) for clinical documentation tasks such as SOAP note generation remains challenging. Unlike standard summarization, these tasks require clinical abstraction, normalization of colloquial language, and medically grounded inference. However, prevailing evaluation methods including automated metrics and LLM as judge frameworks rely on lexical faithfulness, often labeling any information not explicitly present in the transcript as hallucination. We show that such approaches systematically misclassify clinically valid outputs as errors, inflating hallucination rates and distorting model assessment. Our analysis reveals that many flagged hallucinations correspond to legitimate clinical transformations, including synonym mapping, abstraction of examination findings, diagnostic inference, and guideline consistent care planning. By aligning evaluation criteria with clinical reasoning through calibrated prompting and retrieval grounded in medical ontologies we observe a significant shift in outcomes. Under a lexical evaluation regime, the mean hallucination rate is 35%, heavily penalizing valid reasoning. With inference aware evaluation, this drops to 9%, with remaining cases reflecting genuine safety concerns. These findings suggest that current evaluation practices over penalize valid clinical reasoning and may measure artifacts of evaluation design rather than true errors, underscoring the need for clinically informed evaluation in high context domains like medicine.

---

### Pangu-ACE: Adaptive Cascaded Experts for Educational Response Generation on EduBench

Authors: Dinghao Li, Wenlong Zhou, Zhimin Chen, Yuehan Peng, Hong Ni, Chengfu Zou, Guoyu Shi, Yaochen Li
Published: 2026-04-16
Categories: cs.CL
arXiv: https://arxiv.org/abs/2604.14828v1

Educational assistants should spend more computation only when the task needs it. This paper rewrites our earlier draft around the system that was actually implemented and archived in the repository: a sample-level 1B to 7B cascade for the shared-8 EduBench benchmark. The final system, Pangu-ACE, uses a 1B tutor-router to produce a draft answer plus routing signals, then either accepts the draft or escalates the sample to a 7B specialist prompt. We also correct a major offline evaluation bug: earlier summaries over-credited some open-form outputs that only satisfied superficial format checks. After CPU-side rescoring from saved prediction JSONL, the full Chinese test archive (7013 samples) shows that cascade_final improves deterministic quality from 0.457 to 0.538 and format validity from 0.707 to 0.866 over the legacy rule_v2 system while accepting 19.7% of requests directly at 1B. Routing is strongly task dependent: IP is accepted by 1B 78.0% of the time, while QG and EC still escalate almost always. The current archived deployment does not yet show latency gains, so the defensible efficiency story is routing selectivity rather than wall-clock speedup. We also package a reproducible artifact-first paper workflow and clarify the remaining external-baseline gap: GPT-5.4 re-judging is implemented locally, but the configured provider endpoint and key are invalid, so final sampled-baseline alignment with GPT-5.4 remains pending infrastructure repair.

---

### Nautilus: An Auto-Scheduling Tensor Compiler for Efficient Tiled GPU Kernels

Authors: Yifan Zhao, Yuchen Yang, Matei Budiu, Sasa Misailovic
Published: 2026-04-16
Categories: cs.PL, cs.LG
arXiv: https://arxiv.org/abs/2604.14825v1

We present Nautilus, a novel tensor compiler that moves toward fully automated math-to-kernel optimization. Nautilus compiles a high-level algebraic specification of tensor operators into efficient tiled GPU kernels. Nautilus's successive lowering design allows high-level optimizations, expression rewrites, and tile optimizations to be jointly applied in a single end-to-end system. Nautilus presents a novel auto-scheduler that discovers sequences of high-level optimizations, while preserving the regular program structure needed by tile optimizers. Nautilus's auto-scheduler captures complex interactions and trade-offs in the high-level optimizations, including aggressive global transformations like advanced reduction fusion. Nautilus is the first end-to-end tensor compiler capable of starting from a math-like description of attention and automatically discovering FlashAttention-3-like kernels, offloading the entire burden of optimization from the programmer to the compiler. Across five transformer-based models and 150 evaluation configurations on NVIDIA GH200 and RTX 5090 GPUs, Nautilus achieves up to 23% higher throughput than state-of-the-art compilers on GH200 and up to 42% on RTX 5090, while matching or exceeding manually written cuDNN kernels on many long-sequence configurations.

---

### NTIRE 2026 Challenge on Video Saliency Prediction: Methods and Results

Authors: Andrey Moskalenko, Alexey Bryncev, Ivan Kosmynin, Kira Shilovskaya, Mikhail Erofeev, Dmitry Vatolin, Radu Timofte, Kun Wang, Yupeng Hu, Zhiran Li, Hao Liu, Qianlong Xiang et al.
Published: 2026-04-16
Categories: cs.CV, cs.HC, cs.MM
arXiv: https://arxiv.org/abs/2604.14816v1

This paper presents an overview of the NTIRE 2026 Challenge on Video Saliency Prediction. The goal of the challenge participants was to develop automatic saliency map prediction methods for the provided video sequences. The novel dataset of 2,000 diverse videos with an open license was prepared for this challenge. The fixations and corresponding saliency maps were collected using crowdsourced mouse tracking and contain viewing data from over 5,000 assessors. Evaluation was performed on a subset of 800 test videos using generally accepted quality metrics. The challenge attracted over 20 teams making submissions, and 7 teams passed the final phase with code review. All data used in this challenge is made publicly available - https://github.com/msu-video-group/NTIRE26_Saliency_Prediction.

---

### Continued Pretraining of FinBERT on Finnish Histopathological Reports: Train-Time Signals and Proxy Downstream Correlations

Authors: Rami Luisto, Liisa Petäinen, Tommi Grönholm, Jan Böhm, Maarit Ahtiainen, Tomi Lilja, Ilkka Pölönen, Sami Äyrämö
Published: 2026-04-16
Categories: cs.CL
arXiv: https://arxiv.org/abs/2604.14815v2

In Natural Language Processing (NLP) classification tasks where a lack of labeled data is an issue, continued pretraining (CPT) of transformer models on unlabeled data is an established approach. In this paper, we have two aims. (1) We describe our observations from continued pretraining of the Finnish BERT transformer model (FinBERT) on a Finnish histopathological dataset (below, \emph{the Histopathology data}). (2) Since the Histopathology data has no classification labels, we gather public Finnish datasets as proxy data to analyze whether the signals observed in (1) are associated with downstream classification gains. We observe that CPT train-time loss curves differ strongly by domain, and that, in an exploratory analysis, certain CPT-derived features correlate with proxy classification improvement. In particular, this report contributes to the limited literature on NLP for Finnish healthcare data.

---

### Domain Fine-Tuning FinBERT on Finnish Histopathological Reports: Train-Time Signals and Downstream Correlations

Authors: Rami Luisto, Liisa Petäinen, Tommi Grönholm, Jan Böhm, Maarit Ahtiainen, Tomi Lilja, Ilkka Pölönen, Sami Äyrämö
Published: 2026-04-16
Categories: cs.CL
arXiv: https://arxiv.org/abs/2604.14815v1

In NLP classification tasks where little labeled data exists, domain fine-tuning of transformer models on unlabeled data is an established approach. In this paper we have two aims. (1) We describe our observations from fine-tuning the Finnish BERT model on Finnish medical text data. (2) We report on our attempts to predict the benefit of domain-specific pre-training of Finnish BERT from observing the geometry of embedding changes due to domain fine-tuning. Our driving motivation is the common\situation in healthcare AI where we might experience long delays in acquiring datasets, especially with respect to labels.

---

### Learning Ad Hoc Network Dynamics via Graph-Structured World Models

Authors: Can Karacelebi, Yusuf Talha Sahin, Elif Surer, Ertan Onur
Published: 2026-04-16
Categories: cs.LG, cs.MA, cs.NI
arXiv: https://arxiv.org/abs/2604.14811v1

Ad hoc wireless networks exhibit complex, innate and coupled dynamics: node mobility, energy depletion and topology change that are difficult to model analytically. Model-free deep reinforcement learning requires sustained online interaction whereas existing model based approaches use flat state representations that lose per node structure. Therefore we propose G-RSSM, a graph structured recurrent state space model that maintains per node latent states with cross node multi head attention to learn the dynamics jointly from offline trajectories. We apply the proposed method to the downstream task clustering where a cluster head selection policy trains entirely through imagined rollouts in the learned world model. Across 27 evaluation scenarios spanning MANET, VANET, FANET, WSN and tactical networks with N=30 to 1000 nodes, the learned policy maintains high connectivity with only trained for N=50. Herein, we propose the first multi physics graph structured world model applied to combinatorial per node decision making in size agnostic wireless ad hoc networks.

---

### Scalable Model-Based Clustering with Sequential Monte Carlo

Authors: Connie Trojan, Pavel Myshkov, Paul Fearnhead, James Hensman, Tom Minka, Christopher Nemeth
Published: 2026-04-16
Categories: stat.ML, cs.LG, stat.CO
arXiv: https://arxiv.org/abs/2604.14810v1

In online clustering problems, there is often a large amount of uncertainty over possible cluster assignments that cannot be resolved until more data are observed. This difficulty is compounded when clusters follow complex distributions, as is the case with text data. Sequential Monte Carlo (SMC) methods give a natural way of representing and updating this uncertainty over time, but have prohibitive memory requirements for large-scale problems. We propose a novel SMC algorithm that decomposes clustering problems into approximately independent subproblems, allowing a more compact representation of the algorithm state. Our approach is motivated by the knowledge base construction problem, and we show that our method is able to accurately and efficiently solve clustering problems in this setting and others where traditional SMC struggles.

---

### Expert-Guided Class-Conditional Goodness-of-Fit Scores for Interpretable Classification with Informative Missingness: An Application to Seismic Monitoring

Authors: Shahar Cohen, David M. Steinberg, Yael Radzyner, Yochai Ben Horin
Published: 2026-04-16
Categories: stat.ML, cs.LG, stat.AP
arXiv: https://arxiv.org/abs/2604.14809v1

We study a classification problem with three key challenges: pervasive informative missingness, the integration of partial prior expert knowledge into the learning process, and the need for interpretable decision rules. We propose a framework that encodes prior knowledge through an expert-guided class-conditional model for one or more classes, and use this model to construct a small set of interpretable goodness-of-fit features. The features quantify how well the observed data agree with the expert model, isolating the contributions of different aspects of the data, including both observed and missing components. These features are combined with a few transparent auxiliary summaries in a simple discriminative classifier, resulting in a decision rule that is easy to inspect and justify. We develop and apply the framework in the context of seismic monitoring used to assess compliance with the Comprehensive Nuclear-Test-Ban Treaty. We show that the method has strong potential as a transparent screening tool, reducing workload for expert analysts. A simulation designed to isolate the contribution of the proposed framework shows that this interpretable expert-guided method can even outperform strong standard machine-learning classifiers, particularly when training samples are small.

---

### Modeling LLM Unlearning as an Asymmetric Two-Task Learning Problem

Authors: Zeguan Xiao, Siqing Li, Yong Wang, Xuetao Wei, Jian Yang, Yun Chen, Guanhua Chen
Published: 2026-04-16
Categories: cs.CL
arXiv: https://arxiv.org/abs/2604.14808v1

Machine unlearning for large language models (LLMs) aims to remove targeted knowledge while preserving general capability. In this paper, we recast LLM unlearning as an asymmetric two-task problem: retention is the primary objective and forgetting is an auxiliary. From this perspective, we propose a retention-prioritized gradient synthesis framework that decouples task-specific gradient extraction from conflict-aware combination. Instantiating the framework, we adapt established PCGrad to resolve gradient conflicts, and introduce SAGO, a novel retention-prioritized gradient synthesis method. Theoretically, both variants ensure non-negative cosine similarity with the retain gradient, while SAGO achieves strictly tighter alignment through constructive sign-constrained synthesis. Empirically, on WMDP Bio/Cyber and RWKU benchmarks, SAGO consistently pushes the Pareto frontier: e.g., on WMDP Bio (SimNPO+GD), recovery of target model MMLU performance progresses from 44.6% (naive) to 94.0% (+PCGrad) and further to 96.0% (+SAGO), while maintaining comparable forgetting strength. Our results show that re-shaping gradient geometry, rather than re-balancing losses, is the key to mitigating unlearning-retention trade-offs.

---

### The LLM Fallacy: Misattribution in AI-Assisted Cognitive Workflows

Authors: Hyunwoo Kim, Harin Yu, Hanau Yi
Published: 2026-04-16
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2604.14807v2

The rapid integration of large language models (LLMs) into everyday workflows has transformed how individuals perform cognitive tasks such as writing, programming, analysis, and multilingual communication. While prior research has focused on model reliability, hallucination, and user trust calibration, less attention has been given to how LLM usage reshapes users' perceptions of their own capabilities. This paper introduces the LLM fallacy, a cognitive attribution error in which individuals misinterpret LLM-assisted outputs as evidence of their own independent competence, producing a systematic divergence between perceived and actual capability. We argue that the opacity, fluency, and low-friction interaction patterns of LLMs obscure the boundary between human and machine contribution, leading users to infer competence from outputs rather than from the processes that generate them. We situate the LLM fallacy within existing literature on automation bias, cognitive offloading, and human-AI collaboration, while distinguishing it as a form of attributional distortion specific to AI-mediated workflows. We propose a conceptual framework of its underlying mechanisms and a typology of manifestations across computational, linguistic, analytical, and creative domains. Finally, we examine implications for education, hiring, and AI literacy, and outline directions for empirical validation. We also provide a transparent account of human-AI collaborative methodology. This work establishes a foundation for understanding how generative AI systems not only augment cognitive performance but also reshape self-perception and perceived expertise.

---

### From Boundaries to Semantics: Prompt-Guided Multi-Task Learning for Petrographic Thin-section Segmentation

Authors: Yili Ren, Shiqi Wen, Li Hou, Dingwen Xiao, Weiming Zhang, Caleb Chen Cao, Lin Wang, Zilu Zheng, Qianxiao Su, Mingjun Zhao, Lei Chen
Published: 2026-04-16
Categories: cs.CV
arXiv: https://arxiv.org/abs/2604.14805v1

Grain-edge segmentation (GES) and lithology semantic segmentation (LSS) are two pivotal tasks for quantifying rock fabric and composition. However, these two tasks are often treated separately, and the segmentation quality is implausible albeit expensive, time-consuming, and expert-annotated datasets have been used. Recently, foundation models, especially the Segment Anything Model (SAM), have demonstrated impressive robustness for boundary alignment. However, directly adapting SAM to joint GES and LSS is nontrivial due to 1) severe domain gap induced by extinction-dependent color variations and ultra-fine grain boundaries, and 2) lacking novel modules for joint learning on multi-angle petrographic image stacks. In this paper, we propose Petro-SAM, a novel two-stage, multi-task framework that can achieve high-quality joint GES and LSS on petrographic images. Specifically, based on SAM, we introduce a Merge Block to integrate seven polarized views, effectively solving the extinction issue. Moreover, we introduce multi-scale feature fusion and color-entropy priors to refine the detection.

---

### Generative Modeling of Complex-Valued Brain MRI Data

Authors: Marco Schlimbach, Moritz Rempe, Jessica Mnischek, Lukas T. Rotkopf, Jens Weingarten, Jens Kleesiek, Kevin Kröninger
Published: 2026-04-16
Categories: eess.IV, cs.CV, physics.med-ph
arXiv: https://arxiv.org/abs/2604.14800v1

Objective. Standard Magnetic Resonance Imaging (MRI) reconstruction pipelines discard phase information captured during acquisition, despite evidence that it encodes tissue properties relevant to tumor diagnosis. Current machine learning approaches inherit this limitation by operating exclusively on reconstructed magnitude images. The aim of this study is to build a generative framework which is capable of jointly modeling magnitude and phase information of complex-valued MRI scans. Approach. The proposed generative framework combines a conditional variational autoencoder, which compresses complex-valued MRI scans into compact latent representations while preserving phase coherence, with a flow-matching-based generative model. Synthetic sample quality is assessed via a real-versus-synthetic classifier and by training downstream classifiers on synthetic data for abnormal tissue detection. Main results. The autoencoder preserves phase coherence above 0.997. Real-versus-synthetic classification yields low AUROC values between 0.50 and 0.66 across all acquisition sequences, indicating generated samples are nearly indistinguishable from real data. In downstream normal-versus-abnormal classification, classifiers trained entirely on synthetic data achieve an AUROC of 0.880, surpassing the real-data baseline of 0.842 on a publicly available dataset (fastMRI). This advantage persists on an independent external test set from a different institution with biopsy-confirmed labels. Significance. The proposed framework demonstrates the feasibility of jointly modeling magnitude and phase information for normal and abnormal complex-valued brain MRI data. Beyond synthetic data generation, it establishes a foundation for the usage of complete brain MRI information in future diagnostic applications and enables systematic investigation of how magnitude and phase jointly encode pathology-specific features.

---

### Knowing When Not to Answer: Evaluating Abstention in Multimodal Reasoning Systems

Authors: Nishanth Madhusudhan, Vikas Yadav, Alexandre Lacoste
Published: 2026-04-16
Categories: cs.CL, cs.CV
arXiv: https://arxiv.org/abs/2604.14799v1

Effective abstention (EA), recognizing evidence insufficiency and refraining from answering, is critical for reliable multimodal systems. Yet existing evaluation paradigms for vision-language models (VLMs) and multi-agent systems (MAS) assume answerability, pushing models to always respond. Abstention has been studied in text-only settings but remains underexplored multimodally; current benchmarks either ignore unanswerability or rely on coarse methods that miss realistic failure modes. We introduce MM-AQA, a benchmark that constructs unanswerable instances from answerable ones via transformations along two axes: visual modality dependency and evidence sufficiency. Evaluating three frontier VLMs spanning closed and open-source models and two MAS architectures across 2079 samples, we find: (1) under standard prompting, VLMs rarely abstain; even simple confidence baselines outperform this setup, (2) MAS improves abstention but introduces an accuracy-abstention trade-off, (3) sequential designs match or exceed iterative variants, suggesting the bottleneck is miscalibration rather than reasoning depth, and (4) models abstain when image or text evidence is absent, but attempt reconciliation with degraded or contradictory evidence. Effective multimodal abstention requires abstention-aware training rather than better prompting or more agents.

---

### PUFFIN: Protein Unit Discovery with Functional Supervision

Authors: Gökçe Uludoğan, Buse Giledereli, Elif Ozkirimli, Arzucan Özgür
Published: 2026-04-16
Categories: q-bio.BM, cs.LG
arXiv: https://arxiv.org/abs/2604.14796v1

Proteins carry out biological functions through the coordinated action of groups of residues organized into structural arrangements. These arrangements, which we refer to as protein units, exist at an intermediate scale, being larger than individual residues yet smaller than entire proteins. A deeper understanding of protein function can be achieved by identifying these units and their associations with function. However, existing approaches either focus on residue-level signals, rely on curated annotations, or segment protein structures without incorporating functional information, thereby limiting interpretable analysis of structure-function relationships. We introduce PUFFIN, a data-driven framework for discovering protein units by jointly learning structural partitioning and functional supervision. PUFFIN represents proteins as residue-level structure graphs and applies a graph neural network with a structure-aware pooling mechanism that partitions each protein into multi-residue units, with functional supervision that shapes the partition. We show that the learned units are structurally coherent, exhibit organized associations with molecular function, and show meaningful correspondence with curated InterPro annotations. Together, these results demonstrate that PUFFIN provides an interpretable framework for analyzing structure-function relationships using learned protein units and their statistical function associations. We made our source code available at https://github.com/boun-tabi-lifelu/puffin.

---

### Diffusion Crossover: Defining Evolutionary Recombination in Diffusion Models via Noise Sequence Interpolation

Authors: Chisato Kumada, Satoru Hiwa, Tomoyuki Hiroyasu
Published: 2026-04-16
Categories: cs.AI
arXiv: https://arxiv.org/abs/2604.14790v2

Interactive Evolutionary Computation (IEC) provides a powerful framework for optimizing subjective criteria such as human preferences and aesthetics, yet it suffers from a fundamental limitation: in high-dimensional generative representations, defining crossover in a semantically consistent manner is difficult, often leading to a mutation-dominated search. In this work, we explicitly define crossover in diffusion models. We propose Diffusion crossover, which formulates evolutionary recombination as step-wise interpolation of noise sequences in the reverse process of Denoising Diffusion Probabilistic Models (DDPMs). By applying spherical linear interpolation (Slerp) to the noise sequences associated with selected parent images, the proposed method generates offspring that inherit characteristics from both parents while preserving the geometric structure of the diffusion process. Furthermore, controlling the time-step range of interpolation enables a principled trade-off between diversity (exploration) and convergence (exploitation). Experimental results using PCA analysis and perceptual similarity metrics (LPIPS) demonstrate that Diffusion crossover produces perceptually smooth and semantically consistent transitions between parent images. Qualitative interactive evolution experiments further confirm that the proposed method effectively supports human-in-the-loop image exploration. These findings suggest a new perspective: diffusion models are not only powerful generators, but also structured evolutionary search spaces in which recombination can be explicitly defined and controlled.

---

### A Comparative Study of CNN Optimization Methods for Edge AI: Exploring the Role of Early Exits

Authors: Nekane Fernandez, Ivan Valdes, Steven Van Vaerenbergh, Idoia de la Iglesia, Julen Arratibel
Published: 2026-04-16
Categories: cs.AI
arXiv: https://arxiv.org/abs/2604.14789v1

Deploying deep neural networks on edge devices requires balancing accuracy, latency, and resource constraints under realistic execution conditions. To fit models within these constraints, two broad strategies have emerged: static compression techniques such as pruning and quantization, which permanently reduce model size, and dynamic approaches such as early-exit mechanisms, which adapt computational cost at runtime. While both families are widely studied in isolation, they are rarely compared under identical conditions on physical hardware. This paper presents a unified deployment-oriented comparison of static compression and dynamic early-exit mechanisms, evaluated on real edge devices using ONNX based inference pipelines. Our results show that static and dynamic techniques offer fundamentally different trade-offs for edge deployment. While pruning and quantization deliver consistent memory footprint reduction, early-exit mechanisms enable input-adaptive computation savings that static methods cannot match. Their combination proves highly effective, simultaneously reducing inference latency and memory usage with minimal accuracy loss, expanding what is achievable at the edge.

---

### Sequence Search: Automated Sequence Design using Neural Architecture Search

Authors: Rokgi Hong, Hongjun An, Sooyeon Ji, Jongho Lee
Published: 2026-04-16
Categories: cs.AI
arXiv: https://arxiv.org/abs/2604.14788v1

Developing an MR sequence is challenging and remains largely constrained by human intuition. Recently, AI-driven approaches have been proposed; however, most require an initial sequence for parameter optimization or extensive training datasets, limiting their general applicability. In this study, we propose "Sequence Search," an automated sequence design framework based on neural architecture search. The method takes tissue properties, imaging parameters, and design objectives as inputs and generates pulse sequences satisfying the design objectives, without requiring prior knowledge of conventional sequence structures. Sequence Search iteratively generates candidate sequences through neural architecture search and optimizes them via a differentiable Bloch simulator and objective-specific loss functions using gradient-based learning. The framework successfully replicated conventional spin-echo, T2-weighted spin-echo, and inversion recovery sequences. Less intuitive solutions were also discovered, such as three-RF spin-echo-like sequences with reduced RF energy and refocusing phases deviating from the conventional Hahn-echo. This work establishes a generalizable framework for automated MR sequence design, highlighting the potential to explore configurations beyond conventional designs based on human intuition.

---

### Towards Trustworthy 6G Network Digital Twins: A Framework for Validating Counterfactual What-If Analysis in Edge Computing Resources

Authors: Julian Jimenez Agudelo, Paola Soto, Ayat Zaki-Hindi, Jean-Sébastien Sottet, Sébastien Faye, Nina Slamnik-Kriještorac, Johann Marquez-Barja, Miguel Camelo Botero
Published: 2026-04-16
Categories: eess.SY, cs.LG, cs.NI
arXiv: https://arxiv.org/abs/2604.14787v1

Network Digital Twins (NDTs) enable safe what-if analysis for 6G cloud-edge infrastructures, but adoption is often limited by fragmented workflows from telemetry to validation. We present a data-driven NDT framework that extends 6G-TWIN with a scalable pipeline for cloud-edge telemetry aggregation and semantic alignment into unified data models. Our contributions include: (i) scalable cloud-edge telemetry collection, (ii) regime-aware feature engineering capturing the network's scaling behavior, and (iii) a validation methodology based on Sign Agreement and Directional Sensitivity. Evaluated on a Kubernetes-managed cluster, the framework extrapolates performance to unseen high-load regimes. Results show both Deep Neural Network (DNN) and XGBoost achieve high regression accuracy (R2 > 0.99), while the XGBoost model delivers superior directional reliability (Sa > 0.90), making the NDT a trustworthy tool for proactive resource scaling in out-of-distribution scenarios.

---