# Research Papers — 2026-04 · bundle page 21 of 233 > Full-text markdown of papers #1001–1050 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/21 [← Previous bundle](https://aipod.dev/papers/archive/2026-04/page/20.md) [Next bundle →](https://aipod.dev/papers/archive/2026-04/page/22.md) ### G-Loss: Graph-Guided Fine-Tuning of Language Models Authors: Aditya Sharma, Vinti Agarwal, Rajesh Kumar Published: 2026-04-28 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2604.25853v4 Traditional loss functions, including cross-entropy, contrastive, triplet, and su pervised contrastive losses, used for fine-tuning pre-trained language models such as BERT, operate only within local neighborhoods and fail to account for the global semantic structure. We present G-Loss, a graph-guided loss function that incorporates semi-supervised label propagation to use structural relationships within the embedding manifold. G-Loss builds a document-similarity graph that captures global semantic relationships, thereby guiding the model to learn more discriminative and robust embeddings. We evaluate G-Loss on five benchmark datasets covering key downstream classification tasks: MR (sentiment analysis), R8 and R52 (topic categorization), Ohsumed (medical document classification), and 20NG (news categorization). In the majority of experimental setups, G-Loss converges faster and produces semantically coherent embedding spaces, resulting in higher classification accuracy than models fine-tuned with traditional loss functions. --- ### G-Loss: Graph-Guided Fine-Tuning of Language Models Authors: Aditya Sharma, Vinti Agarwal, Rajesh Kumar Published: 2026-04-28 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2604.25853v3 Traditional loss functions, including cross-entropy, contrastive, triplet, and su pervised contrastive losses, used for fine-tuning pre-trained language models such as BERT, operate only within local neighborhoods and fail to account for the global semantic structure. We present G-Loss, a graph-guided loss function that incorporates semi-supervised label propagation to use structural relationships within the embedding manifold. G-Loss builds a document-similarity graph that captures global semantic relationships, thereby guiding the model to learn more discriminative and robust embeddings. We evaluate G-Loss on five benchmark datasets covering key downstream classification tasks: MR (sentiment analysis), R8 and R52 (topic categorization), Ohsumed (medical document classification), and 20NG (news categorization). In the majority of experimental setups, G-Loss converges faster and produces semantically coherent embedding spaces, resulting in higher classification accuracy than models fine-tuned with traditional loss functions. --- ### Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses Authors: Jiahang Lin, Shichun Liu, Chengjun Pan, Lizhi Lin, Shihan Dou, Zhiheng Xi, Xuanjing Huang, Hang Yan, Zhenhua Han, Tao Gui, Yu-Gang Jiang Published: 2026-04-28 Categories: cs.CL, cs.SE arXiv: https://arxiv.org/abs/2604.25850v4 Harnesses are now central to coding-agent performance, mediating how models interact with tools and execution environments. Yet harness engineering remains a manual craft, because automating it faces a heterogeneous action space across editable components, voluminous trajectories that bury actionable signal, and edits whose effect is hard to attribute. We introduce Agentic Harness Engineering (AHE), a closed loop that addresses these challenges through three matched observability pillars: (1) component observability gives every editable harness component a file-level representation so the action space is explicit and revertible; (2) experience observability distills millions of raw trajectory tokens into a layered, drill-down evidence corpus that an evolving agent can actually consume; and (3) decision observability pairs every edit with a self-declared prediction, later verified against the next round's task-level outcomes. Together, these pillars turn every edit into a falsifiable contract, so harness evolution proceeds autonomously without collapsing into trial-and-error. Empirically, ten AHE iterations lift pass@1 on Terminal-Bench 2 from 69.7% to 77.0%, surpassing the human-designed harness Codex-CLI (71.9%) and the self-evolving baselines ACE and TF-GRPO. The frozen harness transfers without re-evolution: on SWE-bench-verified it tops aggregate success at 12% fewer tokens than the seed, and on Terminal-Bench 2 it yields +5.1 to +10.1pp cross-family gains across three alternate model families, indicating the evolved components encode general engineering experience rather than benchmark-specific tuning. Ablations localize the gain to tools, middleware, and long-term memory rather than the system prompt, suggesting factual harness structure transfers while prose-level strategy does not. --- ### ADEMA: A Knowledge-State Orchestration Architecture for Long-Horizon Knowledge Synthesis with LLMAgents Authors: Zhou Hanlin, Chan Huah Yong Published: 2026-04-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2604.25849v1 Long-horizon LLM tasks often fail not because a single answer is unattainable, but because knowledge states drift across rounds, intermediate commitments remain implicit, and interruption fractures the evolving evidence chain. This paper presents ADEMA as a knowledge-state orchestration architecture for long-horizon knowledge synthesis rather than as a generic multi-agent runtime. The architecture combines explicit epistemic bookkeeping, heterogeneous dual-evaluator governance, adaptive task-mode switching, reputation-shaped resource allocation, checkpoint-resumable persistence, segment-level memory condensation, artifact-first assembly, and final-validity checking with safe fallback. Evidence is drawn entirely from existing materials: a four-scenario showcase package, a fixed 60-run mechanism matrix, targeted micro-ablation and artifact-chain supplements, and a repaired protocol-level benchmark in which code-oriented evaluation is the clearest quality-sensitive mechanism block. Across the fixed matrix, removing checkpoint/resume produced the only invalid run, and it did so in the interruption-sensitive resume condition. By contrast, dual evaluation, segment synthesis, and dynamic governance are best interpreted as supporting control mechanisms that shape trajectory discipline, explicit artifact progression, and cost-quality behavior rather than as universal binary prerequisites for completion. The contribution is therefore a knowledge-state orchestration architecture in which explicit epistemic state transition, evidence-bearing artifact progression, and recoverable continuity are the primary design commitments. --- ### A Distributionally Robust Reinforcement Learning Framework for Constrained Urban EV Dispatch Authors: An Nguyen, Hoang Nguyen, Phuong Le, Hung Pham, Cuong Do, Laurent El Ghaoui Published: 2026-04-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2604.25848v2 We study city-scale control of electric-vehicle (EV) ride-hailing fleets where dispatch, repositioning, and charging decisions must respect charger and feeder limits under uncertain, spatially correlated demand and travel times. We formulate the problem as a hex-grid semi-Markov decision process (semi-MDP) with mixed actions -- discrete actions for serving, repositioning, and charging, together with continuous charging power -- and variable action durations. To guarantee physical feasibility during both training and deployment, the policy learns over high-level intentions produced by a masked, temperature-annealed actor. These intentions are projected at every decision step through a time-limited rolling mixed-integer linear program (MILP) that strictly enforces state-of-charge, port, and feeder constraints. To mitigate distributional shifts, we optimize a Soft Actor-Critic (SAC) agent against a Wasserstein-1 ambiguity set with a graph-aligned Mahalanobis ground metric that captures spatial correlations. The robust backup uses the Kantorovich-Rubinstein dual, a projected subgradient inner loop, and a primal-dual risk-budget update. Our architecture combines a two-layer Graph Convolutional Network (GCN) encoder, twin critics, and a value network that drives the adversary. Experiments on a large-scale EV fleet simulator built from NYC taxi data show that PD-RSAC achieves the highest net profit, reaching \$1.22M, compared with \$0.58M-\$0.70M for strong heuristic, single-agent RL, and multi-agent RL baselines, including Greedy, SAC, MAPPO, and MADDPG, while maintaining zero feeder-limit violations. --- ### From Soliloquy to Agora: Memory-Enhanced LLM Agents with Decentralized Debate for Optimization Modeling Authors: Jianghao Lin, Zi Ling, Chenyu Zhou, Tianyi Xu, Ruoqing Jiang, Zizhuo Wang, Dongdong Ge Published: 2026-04-28 Categories: math.OC, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2604.25847v1 Optimization modeling underpins real-world decision-making in logistics, manufacturing, energy, and public services, but reliably solving such problems from natural-language requirements remains challenging for current large language models (LLMs). In this paper, we propose \emph{Agora-Opt}, a modular agentic framework for optimization modeling that combines decentralized debate with a read-write memory bank. Agora-Opt allows multiple agent teams to independently produce end-to-end solutions and reconcile them through an outcome-grounded debate protocol, while memory stores solver-verified artifacts and past disagreement resolutions to support training-free improvement over time. This design is flexible across both backbones and methods: it reduces base-model lock-in, transfers across different LLM families, and can be layered onto existing pipelines with minimal coupling. Across public benchmarks, Agora-Opt achieves the strongest overall performance among all compared methods, outperforming strong zero-shot LLMs, training-centric approaches, and prior agentic baselines. Further analyses show robust gains across backbone choices and component variants, and demonstrate that decentralized debate offers a structural advantage over centralized selection by enabling agents to refine candidate solutions through interaction and even recover correct formulations when all initial candidates are flawed. These results suggest that reliable optimization modeling benefits from combining collaborative cross-checking with reusable experience, and position Agora-Opt as a practical and extensible foundation for trustworthy optimization modeling assistance. Our code and data are available at https://github.com/CHIANGEL/Agora-Opt. --- ### Towards Agentic Investigation of Security Alerts Authors: Even Eilertsen, Vasileios Mavroeidis, Gudmund Grov Published: 2026-04-28 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2604.25846v1 Security analysts are overwhelmed by the volume of alerts and the low context provided by many detection systems. Early-stage investigations typically require manual correlation across multiple log sources, a task that is usually time-consuming. In this paper, we present an experimental, agentic workflow that leverages large language models (LLMs) augmented with predefined queries and constrained tool access (structured SQL over Suricata logs and grep-based text search) to automate the first stages of alert investigation. The proposed workflow integrates queries to provide an overview of the available data, and LLM components that selects which queries to use based on the overview results, extracts raw evidence from the query results, and delivers a final verdict of the alert. Our results demonstrate that the LLM-powered workflow can investigate log sources, plan an investigation, and produce a final verdict that has a significantly higher accuracy than a verdict produced by the same LLM without the proposed workflow. By recognizing the inherent limitations of directly applying LLMs to high-volume and unstructured data, we propose combining existing investigation practices of real-world analysts with a structured approach to leverage LLMs as virtual security analysts, thereby assisting and reducing the manual workload. --- ### Model-agnostic information transfer and fusion for classification with label noise Authors: Zhu Guojun, Zhang Sanguo, Ren Mingyang Published: 2026-04-28 Categories: stat.ME, math.ST, stat.ML arXiv: https://arxiv.org/abs/2604.25845v1 Label noise presents a fundamental challenge in modern machine learning, especially when large-scale datasets are generated via automated processes. An increasingly common and important data paradigm, particularly in domains like medical imaging, involves learning from a large dataset with coarse, noisy labels supplemented by a small, expert-verified, clean dataset. This setting constitutes a typical information transfer and fusion problem. However, the significant distribution shift between the noisy and clean data violates the core overall parametric similarity assumptions of existing statistical transfer learning methods, while their reliance on parametric models is ill-suited for complex data like images. To address these limitations, this paper develops a generic model-agnostic nonparametric framework for classification with label noise, which applies to a broad class of classifiers. Our approach leverages the small clean dataset to ``purify'' the large noisy one and carefully manages the remaining ambiguous samples. This framework is underpinned by a rigorous statistical theory. Its empirical performance is demonstrated through simulations and a real-world application to medical image analysis for pneumonia diagnosis. --- ### PSI-Bench: Towards Clinically Grounded and Interpretable Evaluation of Depression Patient Simulators Authors: Nguyen Khoi Hoang, Shuhaib Mehri, Tse-An Hsu, Yi-Jyun Sun, Quynh Xuan Nguyen Truong, Khoa D Doan, Dilek Hakkani-Tür Published: 2026-04-28 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.25840v1 Patient simulators are gaining traction in mental health training by providing scalable exposure to complex and sensitive patient interactions. Simulating depressed patients is particularly challenging, as safety constraints and high patient variability complicate simulations and underscore the need for simulators that capture diverse and realistic patient behaviors. However, existing evaluations heavily rely on LLM-judges with poorly specified prompts and do not assess behavioral diversity. We introduce PSI-Bench, an automatic evaluation framework that provides interpretable, clinically grounded diagnostics of depression patient simulator behavior across turn-, dialogue-, and population-level dimensions. Using PSI-Bench, we benchmark seven LLMs across two simulator frameworks and find that simulators produce overly long, lexically diverse responses, show reduced variability, resolve emotions too quickly, and follow a uniform negative-to-positive trajectory. We also show that the simulation framework has a larger impact on fidelity than the model scale. Results from a human study demonstrate that our benchmark is strongly aligned with expert judgments. Our work reveals key limitations of current depression patient simulators and provides an interpretable, extensible benchmark to guide future simulator design and evaluation. --- ### Action-Aware Generative Sequence Modeling for Short Video Recommendation Authors: Wenhao Li, Zihan Lin, Zhengxiao Guo, Jie Zhou, Shukai Liu, Yongqi Liu, Chuan Luo, Chaoyi Ma, Ruiming Tang, Han Li Published: 2026-04-28 Categories: cs.AI, cs.IR arXiv: https://arxiv.org/abs/2604.25834v2 With the rapid development of the Internet, users have increasingly higher expectations for the recommendation accuracy of online content consumption platforms. However, short videos often contain diverse segments, and users may not hold the same attitude toward all of them. Traditional binary-classification recommendation models, which treat a video as a single holistic entity, face limitations in accurately capturing such nuanced preferences. Considering that user consumption is a temporal process, this paper demonstrates that the timing of user actions can represent diverse intentions through statistical analysis and examination of action patterns. Based on this insight, we propose a novel modeling paradigm: Action-Aware Generative Sequence Network (A2Gen), which refines user actions along the temporal dimension and chains them into sequences for unified processing and prediction. First, we introduce the Context-aware Attention Module (CAM) to model action sequences enriched with item-specific contextual features. Building upon this, we develop the Hierarchical Sequence Encoder (HSE) to learn temporal action patterns from users' historical actions. Finally, through leveraging CAM, we design a module for action sequence generation: the Action-seq Autoregressive Generator (AAG). Extensive offline experiments on the Kuaishou's dataset and the Tmall public dataset demonstrate the superiority of our proposed model. Furthermore, through large-scale online A/B testing deployed on Kuaishou's platform, our model achieves significant improvements over baseline methods in multi-task prediction by leveraging sequential information. Specifically, it yields increases of 0.34% in user watch time, 8.1% in interaction rate, and 0.162% in overall user retention (LifeTime-7), leading to successful deployment across all traffic, serving over 400 million users every day. --- ### TrialCalibre: A Fully Automated Causal Engine for RCT Benchmarking and Observational Trial Calibration Authors: Amir Habibdoust, Xing Song Published: 2026-04-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2604.25832v1 Real-world evidence (RWE) studies that emulate target trials increasingly inform regulatory and clinical decisions, yet residual, hard-to-quantify biases still limit their credibility. The recently proposed BenchExCal framework addresses this challenge via a two-stage Benchmark, Expand, Calibrate process, which first compares an observational emulation against an existing randomized controlled trial (RCT), then uses observed divergence to calibrate a second emulation for a new indication causal effect estimation. While methodologically powerful, BenchExCal is resource intensive and difficult to scale. We introduce TrialCalibre, a conceptualized multiagent system designed to automate and scale the BenchExCal workflow. Our framework features specialized agents such as the Orchestrator, Protocol Design, Data Synthesis, Clinical Validation, and Quantitative Calibration Agents that coordi-nate the the overall process. TrialCalibre incorpo-rates agent learning (e.g., RLHF) and knowledge blackboards to support adaptive, auditable, and transparent causal effect estimation. --- ### Mutual Forcing: Dual-Mode Self-Evolution for Fast Autoregressive Audio-Video Character Generation Authors: Yupeng Zhou, Lianghua Huang, Zhifan Wu, Jiabao Wang, Yupeng Shi, Biao Jiang, Daquan Zhou, Yu Liu, Ming-Ming Cheng, Qibin Hou Published: 2026-04-28 Categories: cs.CV, cs.SD arXiv: https://arxiv.org/abs/2604.25819v1 In this work, we propose Mutual Forcing, a framework for fast autoregressive audio-video generation with long-horizon audio-video synchronization. Our approach addresses two key challenges: joint audio-video modeling and fast autoregressive generation. To ease joint audio-video optimization, we adopt a two-stage training strategy: we first train uni-modal generators and then couple them into a unified audio-video model for joint training on paired data. For streaming generation, we ask whether a native fast causal audio-video model can be trained directly, instead of following existing streaming distillation pipelines that typically train a bidirectional model first and then convert it into a causal generator through multiple distillation stages. Our answer is Mutual Forcing, which builds directly on native autoregressive model and integrates few-step and multi-step generation within a single weight-shared model, enabling self-distillation and improved training-inference consistency. The multi-step mode improves the few-step mode via self-distillation, while the few-step mode generates historical context during training to improve training-inference consistency; because the two modes share parameters, these two effects reinforce each other within a single model. Compared with prior approaches such as Self-Forcing, Mutual Forcing removes the need for an additional bidirectional teacher model, supports more flexible training sequence lengths, reduces training overhead, and allows the model to improve directly from real paired data rather than a fixed teacher. Experiments show that Mutual Forcing matches or surpasses strong baselines that require around 50 sampling steps while using only 4 to 8 steps, demonstrating substantial advantages in both efficiency and quality. The project page is available at https://mutualforcing.github.io. --- ### Magnification-Invariant Image Classification via Domain Generalization and Stable Sparse Embedding Signatures Authors: Ifeanyi Ezuma, Olusiji Medaiyese Published: 2026-04-28 Categories: cs.CV, stat.ML arXiv: https://arxiv.org/abs/2604.25817v1 Magnification shift is a major obstacle to robust histopathology classification, because models trained on one imaging scale often generalize poorly to another. Here, we evaluated this problem on the BreaKHis dataset using a strict patient-disjoint leave-one-magnification-out protocol, comparing supervised baseline, baseline augmented with DCGAN-generated patches, and a gradient-reversal domain-general model designed to preserve discriminative information while suppressing magnification-specific variation. Across held-out magnifications, the domain-general model achieved the strongest overall discrimination and its clearest gain was observed when 200X was held out. By contrast, GAN augmentation produced inconsistent effects, improving some folds but degrading others, particularly at 400X. The domain-general model also yielded the lowest Brier score at 0.063 vs 0.089 at baseline. Sparse embedding analysis further revealed that domain-general training reduced average signature size more than three-fold (306 versus 1,074 dimensions) while preserving equivalent predictive performance (AUC: 0.967 vs 0.965; F1: 0.930 vs 0.931). It also increased cross-fold signature reproducibility from near-zero Jaccard overlap in the baseline to 0.99 between the 100X and 200X folds. These findings show that calibrated, compact, and transferable representations can be learned without added architectural complexity, with clear implications for the reliable deployment of computational pathology models across heterogeneous acquisition settings. --- ### Instruction-Evidence Contrastive Dual-Stream Decoding for Grounded Vision-Language Reasoning Authors: Yashwant Pravinrao Bangde, Debaditya Roy Published: 2026-04-28 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.25809v2 Vision-Language Models (VLMs) exhibit strong performance in instruction following and open-ended vision-language reasoning, yet they frequently generate fluent outputs that are weakly grounded in visual evidence. Prior works have shown that instruction prompting further worsens this issue by amplifying language priors, especially when the visual signal is uncertain or ambiguous. To address this challenge, we propose a decoding framework that explicitly balances linguistic informativeness and visual faithfulness during generation. Our method, Instruction-Evidence Contrastive Dual-Stream Decoding (IECD$^2$), maintains two parallel probability distribution of tokens at each decoding step: an instruction-driven stream that promotes expressive and informative responses, and an evidence-driven stream that enforces strict grounding in the image. These two streams are adaptively fused using a symmetric KL-based contrastive gate, which suppresses tokens favored by language priors but unsupported by visual evidence, while preserving them when both distributions agree. We evaluate IECD$^2$ on multiple datasets spanning various generative vision-language reasoning tasks such as captioning and visual question answering on multiple datasets such as, POPE, MME, VQAv2, AMBER, and MSCOCO. IECD$^2$ demonstrates consistent improvements in task accuracy and reasoning performance with substantial reduction in hallucination compared to state-of-the-art decoding approaches. --- ### MAIC-UI: Making Interactive Courseware with Generative UI Authors: Shangqing Tu, Yanjia Li, Keyu Chen, Sichen Zhang, Jifan Yu, Daniel Zhang-Li, Lei Hou, Juanzi Li, Yu Zhang, Huiqin Liu Published: 2026-04-28 Categories: cs.CL, cs.AI, cs.HC arXiv: https://arxiv.org/abs/2604.25806v1 Creating interactive STEM courseware traditionally requires HTML/CSS/JavaScript expertise, leaving barriers for educators. While generative AI can produce HTML codes, existing tools generate static presentations rather than interactive simulations, struggle with long documents, and lack pedagogical accuracy mechanisms. Furthermore, full regeneration for modifications requires 200--600 seconds, disrupting creative flow. We present MAIC-UI, a zero-code authoring system that enables educators to create and rapidly edit interactive courseware from textbooks, PPTs, and PDFs. MAIC-UI employs: (1) structured knowledge analysis with multi-modal understanding to ensure pedagogical rigor; (2) a two-stage generate-verify-optimize pipeline separating content alignment from visual refinement; and (3) Click-to-Locate editing with Unified Diff-based incremental generation achieving sub-10-second iteration cycles. A controlled lab study with 40 participants shows MAIC-UI reduces editing iterations (4.9 vs. 7.0) and significantly improves learnability and controllability compared to direct Text-to-HTML generation. A three-month classroom deployment with 53 high school students demonstrates that MAIC-UI fosters learning agency and reduces outcome disparities -- the pilot class achieved 9.21-point gains in STEM subjects compared to -2.32 points in control classes. Our code is available at https://github.com/THU-MAIC/MAIC-UI. --- ### Barriers to Universal Reasoning With Transformers (And How to Overcome Them) Authors: Oliver Kraus, Yash Sarrof, Yuekun Yao, Alexander Koller, Michael Hahn Published: 2026-04-28 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2604.25800v2 Chain-of-Thought (CoT) has been shown to empirically improve Transformers' performance, and theoretically increase their expressivity to Turing completeness. However, whether Transformers can learn to generalize to CoT traces longer than those seen during training is understudied. We use recent theoretical frameworks for Transformer length generalization and find that -- under standard positional encodings and a finite alphabet -- Transformers with CoT cannot solve problems beyond $TC^0$, i.e. the expressivity benefits do not hold under the stricter requirement of length-generalizable learnability. However, if we allow the vocabulary to grow with problem size, we attain a length-generalizable simulation of Turing machines where the CoT trace length is linear in the simulated runtime up to a constant. Our construction overcomes two core obstacles to reliable length generalization: repeated copying and last-occurrence retrieval. We assign each tape position a unique signpost token, and log only value changes to enable recovery of the current tape symbol through counts circumventing both barriers. Further, we empirically show that the use of such signpost tokens and value change encodings provide actionable guidance to improve length generalization on hard problems. --- ### Barriers to Universal Reasoning With Transformers (And How to Overcome Them) Authors: Oliver Kraus, Yash Sarrof, Yuekun Yao, Alexander Koller, Michael Hahn Published: 2026-04-28 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2604.25800v1 Chain-of-Thought (CoT) has been shown to empirically improve Transformers' performance, and theoretically increase their expressivity to Turing completeness. However, whether Transformers can learn to generalize to CoT traces longer than those seen during training is understudied. We use recent theoretical frameworks for Transformer length generalization and find that -- under standard positional encodings and a finite alphabet -- Transformers with CoT cannot solve problems beyond $TC^0$, i.e. the expressivity benefits do not hold under the stricter requirement of length-generalizable learnability. However, if we allow the vocabulary to grow with problem size, we attain a length-generalizable simulation of Turing machines where the CoT trace length is linear in the simulated runtime up to a constant. Our construction overcomes two core obstacles to reliable length generalization: repeated copying and last-occurrence retrieval. We assign each tape position a unique signpost token, and log only value changes to enable recovery of the current tape symbol through counts circumventing both barriers. Further, we empirically show that the use of such signpost tokens and value change encodings provide actionable guidance to improve length generalization on hard problems. --- ### At the Edge of the Heart: ULP FPGA-Based CNN for On-Device Cardiac Feature Extraction in Smart Health Sensors for Astronauts Authors: Kazi Mohammad Abidur Rahman, Davis Rakhshan, Philipp Lütke, Laura Harms, Ulf Kulau Published: 2026-04-28 Categories: cs.AR, cs.AI arXiv: https://arxiv.org/abs/2604.25799v1 The convergence of accelerating human spaceflight ambitions and critical terrestrial health monitoring demands is driving unprecedented requirements for reliable, real-time feature extraction on extremely resource-constrained wearable health sensors. We present an ultra-low-power (ULP) Field-Programmable Gate Array (FPGA) based solution for real-time Seismocardiography (SCG) feature classification using Convolutional Neural Networks (CNNs). Our approach combines quantization-aware training with a systolic-array accelerator to enable efficient integer-only inference on the Lattice iCE40UP5K FPGA, which offers an ideal platform for battery-powered deployments -- particularly in space environments -- thanks to its power efficiency and radiation resilience. The implementation achieves a validation accuracy of 98% while consuming only 8.55 mW, completing inference in 95.5 ms with minimal hardware resources (2,861 LUTs and 7 DSP blocks). These results demonstrate that fully on-device SCG-based cardiac feature extraction is feasible on resource-constrained hardware, enabling energy-efficient, autonomous health monitoring for astronauts in long-duration space missions. --- ### StratFormer: Adaptive Opponent Modeling and Exploitation in Imperfect-Information Games Authors: Andy Caen, Mark H. M. Winands, Dennis J. N. J. Soemers Published: 2026-04-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2604.25796v1 We present StratFormer, a transformer-based meta-agent that learns to simultaneously model and exploit opponents in imperfect-information games through a two-phase curriculum. The first phase trains an opponent modeling head to identify behavioral patterns from action histories while the agent plays a game-theoretic optimal (GTO) policy. The second phase progressively shifts the policy toward best-response (BR) exploitation, guided by a per-opponent regularization schedule tied to exploitability. Our architecture introduces dual-turn tokens -- feature vectors constructed at both agent and opponent decision points -- coupled with bucket-rate features that encode opponent tendencies across five strategic contexts. On Leduc Hold'em, a small poker variant with six cards and two betting rounds, we test against six opponent archetypes at two strength levels each, with exploitability ranging from 0.15 to 1.26 Big Blinds (BB) per hand. StratFormer achieves an average exploitation gain of +0.106 BB per hand over GTO, with peak gains of +0.821 against highly exploitable opponents, while maintaining near-equilibrium safety. --- ### Improving Diversity in Black-box Few-shot Knowledge Distillation Authors: Tri-Nhan Vo, Dang Nguyen, Kien Do, Sunil Gupta Published: 2026-04-28 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2604.25795v1 Knowledge distillation (KD) is a well-known technique to effectively compress a large network (teacher) to a smaller network (student) with little sacrifice in performance. However, most KD methods require a large training set and internal access to the teacher, which are rarely available due to various restrictions. These challenges have originated a more practical setting known as black-box few-shot KD, where the student is trained with few images and a black-box teacher. Recent approaches typically generate additional synthetic images but lack an active strategy to promote their diversity, a crucial factor for student learning. To address these problems, we propose a novel training scheme for generative adversarial networks, where we adaptively select high-confidence images under the teacher's supervision and introduce them to the adversarial learning on-the-fly. Our approach helps expand and improve the diversity of the distillation set, significantly boosting student accuracy. Through extensive experiments, we achieve state-of-the-art results among other few-shot KD methods on seven image datasets. The code is available at https://github.com/votrinhan88/divbfkd. --- ### Diverse Image Priors for Black-box Data-free Knowledge Distillation Authors: Tri-Nhan Vo, Dang Nguyen, Trung Le, Kien Do, Sunil Gupta Published: 2026-04-28 Categories: cs.LG, cs.CV arXiv: https://arxiv.org/abs/2604.25794v1 Knowledge distillation (KD) represents a vital mechanism to transfer expertise from complex teacher networks to efficient student models. However, in decentralized or secure AI ecosystems, privacy regulations and proprietary interests often restrict access to the teacher's interface and original datasets. These constraints define a challenging black-box data-free KD scenario where only top-1 predictions and no training data are available. While recent approaches utilize synthetic data, they still face limitations in data diversity and distillation signals. We propose Diverse Image Priors Knowledge Distillation (DIP-KD), a framework that addresses these challenges through a three-phase collaborative pipeline: (1) Synthesis of image priors to capture diverse visual patterns and semantics; (2) Contrast to enhance the collective distinction between synthetic samples via contrastive learning; and (3) Distillation via a novel primer student that enables soft-probability KD. Our evaluation across 12 benchmarks shows that DIP-KD achieves state-of-the-art performance, with ablations confirming data diversity as critical for knowledge acquisition in restricted AI environments. --- ### Subliminal Steering: Stronger Encoding of Hidden Signals Authors: George Morgulis, John Hewitt Published: 2026-04-28 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.25783v2 Subliminal learning describes a student language model inheriting a behavioral bias by fine-tuning on seemingly innocuous data generated by a biased teacher model. Prior work has begun to characterize this phenomenon but leaves open questions about the scope of signals it can transfer, the mechanisms that explain it, and the precision with which a bias can be encoded. We tackle these problems by introducing subliminal steering, a variant of subliminal learning in which the teacher's bias is implemented not via a system prompt, as in prior work, but through a steering vector trained to maximize the likelihood of a set of target samples. First, we show that subliminal steering transfers complex multi-word biases, whereas prior work focused on single-word preferences, demonstrating a large scope of subliminally transferable signals. Moreover, the transfer is reliable enough to appear in settings previously thought not to exhibit subliminal learning, including plain SGD (not just Adam), full fine-tuning (not just LoRA), and models such as Llama and Phi. Second, we provide mechanistic evidence that subliminal learning transfers not only the target behavioral bias, but also the steering vector itself, localized to the layers at which the teacher was steered. Finally, we show that the bias is encoded with such precision that a new steering vector trained on the subliminally-laden dataset attains high cosine similarity with the original vector. --- ### Subliminal Steering: Stronger Encoding of Hidden Signals Authors: George Morgulis, John Hewitt Published: 2026-04-28 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.25783v1 Subliminal learning describes a student language model inheriting a behavioral bias by fine-tuning on seemingly innocuous data generated by a biased teacher model. Prior work has begun to characterize this phenomenon but leaves open questions about the scope of signals it can transfer, the mechanisms that explain it, and the precision with which a bias can be encoded by seemingly unrelated data. We tackle all three problems by introducing subliminal steering, a variant of subliminal learning in which the teacher's bias is implemented not via a system prompt, as in prior work, but through a steering vector trained to maximize the likelihood of a set of target samples. First, we show that subliminal steering transfers complex multi-word biases, whereas prior work focused on single-word preferences, demonstrating a large scope of subliminally transferrable signals. Second, we provide mechanistic evidence that subliminal learning transfers not only the target behavioral bias, but also the steering vector itself, localized to the layers at which the teacher was steered. Finally, we show that the bias is encoded with surprising precision. We train a new steering vector directly on the subliminally-laden dataset and find that it attains high cosine similarity with the original vector. --- ### Sketch2Arti: Sketch-based Articulation Modeling of CAD Objects Authors: Yi Yang, Hao Pan, Yijing Cui, Alla Sheffer, Changjian Li Published: 2026-04-28 Categories: cs.CV, cs.GR arXiv: https://arxiv.org/abs/2604.25781v3 Articulation modeling aims to infer movable parts and their motion parameters for a 3D object, enabling interactive animation, simulation, and shape editing. In this paper, we present Sketch2Arti, the first sketch-based articulation modeling system for CAD objects. Our key observation is that designers naturally communicate articulation intent through lightweight sketches (e.g., arrows and strokes) that indicate how parts should move, yet translating such sketches into articulated 3D models remains largely manual. Sketch2Arti bridges this gap by enabling users to specify articulation through simple 2D sketches drawn from a chosen viewpoint. Given a CAD model and user sketches, our approach automatically discovers the corresponding movable parts and predicts their motion parameters, allowing iterative modeling of multiple articulations on complex objects with fine-grained control. Importantly, Sketch2Arti is trained in a category-agnostic manner without requiring object category information, leading to strong generalization to diverse objects beyond existing articulation datasets. Moreover, for shell models lacking interior structures, Sketch2Arti supports controllable internal completion guided by user sketches, generating plausible internal components consistent with the existing geometry and predicted motion constraints. Comprehensive experiments and user evaluations demonstrate the effectiveness, controllability, and generalization of Sketch2Arti. The code, dataset, and the prototype system are at https://arlo-yang.github.io/Sketch2Arti. --- ### Sustained Gradient Alignment Mediates Subliminal Learning in a Multi-Step Setting: Evidence from MNIST Auxiliary Logit Distillation Experiment Authors: Chayanon Kitkana, Shivam Arora Published: 2026-04-28 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2604.25779v1 In the MNIST auxiliary logit distillation experiment, a student can acquire an unintended teacher trait despite distilling only on no-class logits through a phenomenon called subliminal learning. Under a single-step gradient descent assumption, subliminal learning theory attributes this effect to alignment between the trait and distillation gradients, but does not guarantee that this alignment persists in a multi-step setting. We empirically show that gradient alignment remains weakly but consistently positive throughout training and causally contributes to trait acquisition. We show that a mitigation method called liminal training works by attenuating the alignment and fails to stop trait acquisition in this setup. These results suggest that mitigation methods that operate in this regime may not reliably suppress trait acquisition when the first-order drive dominates. --- ### Can Code Evaluation Metrics Detect Code Plagiarism? Authors: Fahad Ebrahim, Mike Joy Published: 2026-04-28 Categories: cs.SE, cs.AI, cs.IR arXiv: https://arxiv.org/abs/2604.25778v1 Source Code Plagiarism Detection (SCPD) plays an important role in maintaining fairness and academic integrity in software engineering education. Code Evaluation Metrics (CEMs) are developed for assessing code generation tasks. However, it remains unclear whether such metrics can reliably detect plagiarism across different levels of modification (L1-L6), increasing in complexity. In this paper, we perform a comparative empirical study using two open-source labelled datasets, ConPlag (raw and template-free versions) and IRPlag. We evaluate five CEMs, namely CodeBLEU, CrystalBLEU, RUBY, Tree Structured Edit Distance (TSED), and CodeBERTScore. The performance is evaluated using threshold-free ranking-based measures to assess overall, per dataset, and per-level plagiarism performance. The results are compared against state-of-the-art (SOTA) Source Code Plagiarism Detection Tools (SCPDTs), JPlag and Dolos. Our findings show that without preprocessing, Dolos achieves the highest overall ranking performance, while among the individual metrics, CrystalBLEU, CodeBLEU, and RUBY outperform JPlag. Performance is strongest at L1 and drops from L4 onward, while CrystalBLEU remains competitive on L6. With preprocessing, CrystalBLEU surpasses Dolos overall. Per dataset, Dolos achieved the best ranking on the ConPlag raw dataset, while CrystalBLEU was the best-performing metric on the remaining datasets. At the plagiarism levels, Dolos remains strongest on L4, while Crystal-BLEU leads most of the remaining difficult levels. These results indicate that CEMs are comparable to dedicated tools in terms of ranking metrics. --- ### Unrequited Emotions: Investigating the Gaps in Motivation and Practice in Speech Emotion Recognition Research Authors: Taryn Wong, Zeerak Talat, Hanan Aldarmaki, Anjalie Field Published: 2026-04-28 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.25776v1 Critical analyses of emotion recognition technology have raised ethical concerns around task validity and potential downstream impacts, urging researchers to ensure alignment between their stated motivations and practice. However, these discussions have not adequately influenced or drawn from research on speech emotion recognition (SER). We address this gap by conducting a systematic survey of SER research to uncover what stated motivations drive this work and if they align with the datasets and emotions studied. We find that while SER research identifies appealing goals, such as well-situated voice-activated systems or healthcare applications, commonly-used datasets do not reflect these proposed deployment contexts, thus presenting a gap between motivations and research practices. We argue that such gaps engender ethical concerns, and that SER research should reassert itself with concrete use-cases to prevent misinterpretations, misuse, and downstream harms. --- ### CGU-ILALab at FoodBench-QA 2026: Comparing Traditional and LLM-based Approaches for Recipe Nutrient Estimation Authors: Wei-Chun Chen, Yu-Xuan Chen, I-Fang Chung, Ying-Jia Lin Published: 2026-04-28 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.25774v1 Accurate nutrient estimation from unstructured recipe text is an important yet challenging problem in dietary monitoring, due to ambiguous ingredient terminology and highly variable quantity expressions. We systematically evaluate models spanning a wide range of representational capacity, from lexical matching methods (TF-IDF with Ridge Regression), to deep semantic encoders (DeBERTa-v3), to generative reasoning with large language models (LLMs). Under the strict tolerance criteria defined by EU Regulation 1169/2011, our empirical results reveal a clear trade-off between predictive accuracy and computational efficiency. The TF-IDF baseline achieves moderate nutrient estimation performance with near-instantaneous inference, whereas the DeBERTa-v3 encoder performs poorly under task-specific data scarcity. In contrast, few-shot LLM inference (e.g., Gemini 2.5 Flash) and a hybrid LLM refinement pipeline (TF-IDF combined with Gemini 2.5 Flash) deliver the highest validation accuracy across all nutrient categories. These improvements likely arise from the ability of LLMs to leverage pre-trained world knowledge to resolve ambiguous terminology and normalize non-standard units, which remain difficult for purely lexical approaches. However, these gains come at the cost of substantially higher inference latency, highlighting a practical deployment trade-off between real-time efficiency and nutritional precision in dietary monitoring systems. --- ### Measuring the Sensitivity of Classification Models with the Error Sensitivity Profile Authors: Andrea Maurino Published: 2026-04-28 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2604.25765v1 The quality of training data is critical to the performance of machine learning models. In this paper, the Error Sensitivity Profile (ESP) is proposed. It quantifies the sensitivity of model performance to errors in a single feature or in multiple features. By leveraging ESP, data-cleaning efforts can be prioritized based on error types and features most likely to affect model performance. To support the computation of this metric, an integrated suite of tools, called \dirty, is created. We conduct an extensive experimental study on two widely used datasets using 14 classification models, revealing that performance degradation is not always predictable from simple correlations with the target variable. --- ### Threat-Oriented Digital Twinning for Security Evaluation of Autonomous Platforms Authors: Thomas J. Neubert, Laxima Niure Kandel, Berker Peköz Published: 2026-04-28 Categories: cs.CR, cs.AI, cs.RO, eess.SY arXiv: https://arxiv.org/abs/2604.25757v1 Open, unclassified research on secure autonomy is constrained by limited access to operational platforms, contested communications infrastructure, and representative adversarial test conditions. This paper presents a threat-oriented digital twinning methodology for cybersecurity evaluation of learning-enabled autonomous platforms. The approach is instantiated as an open-source, modular twin of a representative autonomy stack with separated sensing, autonomy, and supervisory-control functions; confidence-gated multi-modal perception; explicit command and telemetry trust boundaries; and runtime hold-safe behavior. The contribution is methodological: a reproducible design pattern that translates threat analysis into observable, controllable tests for spoofing, replay, malformed-input injection, degraded sensing, and adversarial ML stress. Although the implemented proxy is ground based, the architecture is intentionally framed around stack elements shared with UAV and space systems, including constrained onboard compute, intermittent or high-latency links, probabilistic perception, and mission-critical recovery behavior. The result is an implementable research scaffold for dependable and secure autonomy studies across UAV and space domains. --- ### Quantum-Inspired Robust and Scalable SAR Object Classification Authors: Maximilian Scharf, Marco Trenti, Felix Bock, Padraig Davidson, Tobias Brosch, Benjamin Rodrigues de Miranda, Sigurd Huber, Timo Felser Published: 2026-04-28 Categories: quant-ph, cs.CV, physics.comp-ph arXiv: https://arxiv.org/abs/2604.25755v2 SAR image classification naturally has to deal with huge noise and a high dynamic range particularly requiring robust classification models. Additionally, the deployment of these models on edge devices, such as drones and military aircraft, requires a careful balance between model size and classification accuracy. This study explores the potential of tensor networks to meet these robustness requirements, specifically evaluating their resilience to data poisoning. Unlike previous works that concentrated on conventional neural networks for SAR object detection, this research focuses on the robustness and model reduction capabilities of tensor networks in object classification. Our findings indicate that tensor networks are adept at addressing both the challenges of robustness and the need for model efficiency, thereby contributing valuable insights to the ongoing discourse in radar applications and deep learning methodologies in general. --- ### QAROO: AI-Driven Online Task Offloading for Energy-Efficient and Sustainable MEC Networks Authors: Yongtao Yao, Yao Yang, Haorui Shi, Canglu Zhu, Miaojiang Chen, Ahmed Farouk Published: 2026-04-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2604.25740v1 With the rapid advancement of artificial intelligence (AI) and intelligent science, intelligent edge computing has been widely adopted. However, the limitations of traditional methods, such as poor adaptability and the slow convergence of heuristic algorithms, are becoming increasingly evident. To enable sustainable and resource-efficient edge applications, this paper proposes an online task offloading framework for wireless powered mobile edge computing (MEC) networks, called Quantum Attention-based Reinforcement learning for Online Offloading (QAROO). The system employs a binary offloading strategy with the aim of co-optimizing computing and energy resources in dynamic channel environments. In response to the issues of poor adaptability in traditional approaches and the slow convergence of heuristic algorithms, the framework integrates quantum neural networks and attention mechanisms, introducing three key improvements: using recurrent neural networks to enhance temporal modeling capability, proposing an uncertainty-guided quantization method to improve exploration efficiency, and incorporating attention mechanisms into quantum networks to strengthen feature representation. Experiments demonstrate that the proposed method outperforms comparative schemes in terms of normalized computation speed and processing time, offering an efficient and stable solution for online task offloading in large-scale Internet of Things (IoT) dynamic environments. --- ### SAFEdit: Does Multi-Agent Decomposition Resolve the Reliability Challenges of Instructed Code Editing? Authors: Noam Tarshish, Nofar Selouk, Daniel Hodisan, Bar Ezra Gafniel, Yuval Elovici, Asaf Shabtai, Eliya Nachmani Published: 2026-04-28 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2604.25737v1 Instructed code editing is a significant challenge for large language models (LLMs). On the EditBench benchmark, 39 of 40 evaluated models obtain a task success rate (TSR) below 60 percent, highlighting a gap between general code generation and the ability to perform instruction-driven editing under executable test constraints. To address this, we propose SAFEdit, a multi-agent framework for instructed code editing that decomposes the editing process into specialized roles to improve reliability and reduce unintended code changes. A Planner Agent produces an explicit, visibility-aware edit plan, an Editor Agent applies minimal, literal code modifications, and a Verifier Agent executes real test runs. When tests fail, SAFEdit uses a Failure Abstraction Layer (FAL) to transform raw test logs into structured diagnostic feedback, which is fed back to the Editor to support iterative refinement. We compare SAFEdit against both prior single-model results reported for EditBench and an implemented ReAct single-agent baseline under the same evaluation conditions. We used EditBench to evaluate SAFEdit on 445 code editing instances in five languages (English, Polish, Spanish, Chinese, and Russian) under varying spatial context variants. SAFEdit achieved 68.6 percent TSR, outperforming the single-model baseline by 3.8 percentage points and the ReAct single-agent baseline by 8.6 percentage points. The iterative refinement loop was found to contribute 17.4 percentage points to SAFEdit's overall success rate. SAFEdit's automated error analysis further indicates a reduction in instruction-level hallucinations compared to single-agent approaches, providing an additional framework component for interpreting failures beyond pass or fail outcomes. --- ### Verification of Neural Networks (Lecture Notes) Authors: Benedikt Bollig Published: 2026-04-28 Categories: cs.LO, cs.AI, cs.FL arXiv: https://arxiv.org/abs/2604.25733v1 These lecture notes provide an introduction to the verification of neural networks from a theoretical perspective. We discuss feed-forward neural networks, recurrent neural networks, attention mechanisms, and transformers, together with specification languages and algorithmic verification techniques. --- ### Toward Scalable Terminal Task Synthesis via Skill Graphs Authors: Zhiyuan Fan, Tinghao Yu, Yuanjun Cai, Jiangtao Guan, Yun Yang, Dingxin Hu, Jiang Zhou, Xing Wu, Zhuo Han, Feng Zhang, Lilin Wang Published: 2026-04-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2604.25727v1 Terminal agents have demonstrated strong potential for autonomous command-line execution, yet their training remains constrained by the scarcity of high-quality and diverse execution trajectories. Existing approaches mitigate this bottleneck by synthesizing large-scale terminal task instances for trajectory sampling. However, they primarily focus on scaling the number of tasks while providing limited control over the diversity of execution trajectories that agents actually experience during training. In this paper, we present SkillSynth, an automated framework for terminal task synthesis built on a scenario-mediated skill graph. SkillSynth first constructs a large-scale skill graph, where scenarios serve as intermediate transition nodes that connect diverse command-line skills. It then samples paths from this graph as abstractions of real-world workflows, and uses a multi-agent harness to instantiate them into executable task instances. By grounding task synthesis in graph-sampled workflow paths, SkillSynth explicitly controls the diversity of minimal execution trajectories required to solve the synthesized tasks. Experiments on Terminal-Bench demonstrate the effectiveness of SkillSynth. Moreover, task instances synthesized by SkillSynth have been adopted to train Hy3 Preview, contributing to its enhanced agentic capabilities in terminal-based settings. --- ### Scalable Inference Architectures for Compound AI Systems: A Production Deployment Study Authors: Srikanta Prasad S, Utkarsh Arora Published: 2026-04-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2604.25724v1 Modern enterprise AI applications increasingly rely on compound AI systems - architectures that compose multiple models, retrievers, and tools to accomplish complex tasks. Deploying such systems in production demands inference infrastructure that can efficiently serve concurrent, heterogeneous model invocations while maintaining cost-effectiveness and low latency. This paper presents a production deployment study of a modular, platform-agnostic inference architecture developed at Salesforce to support compound AI use cases including Agentforce (autonomous AI agents) and ApexGuru (AI-powered code analysis). The system integrates serverless execution, dynamic autoscaling, and MLOps pipelines to deliver consistent low-latency inference across multi-component agent workflows. We report production results demonstrating over 50% reduction in tail latency (P95), up to 3.9x throughput improvement, and 30 to 40% cost savings compared to prior static deployments. We further present a novel analysis of compound-system-specific challenges including multi-model fan-out overhead, cascading cold-start propagation, and heterogeneous scaling dynamics that emerge uniquely when serving agentic workloads. Through detailed case studies and operational lessons, we illustrate how the architecture enables compound AI systems to scale model invocations in parallel, handle bursty multi-agent workloads, and support rapid model iteration - capabilities essential for operationalizing agentic AI at enterprise scale. --- ### Toward Multimodal Conversational AI for Age-Related Macular Degeneration Authors: Ran Gu, Benjamin Hou, Mélanie Hébert, Asmita Indurkar, Yifan Yang, Emily Y. Chew, Tiarnán D. L. Keenan, Zhiyong Lu Published: 2026-04-28 Categories: cs.CV, cs.CL arXiv: https://arxiv.org/abs/2604.25720v1 Despite strong performance of deep learning models in retinal disease detection, most systems produce static predictions without clinical reasoning or interactive explanation. Recent advances in multimodal large language models (MLLMs) integrate diagnostic predictions with clinically meaningful dialogue to support clinical decision-making and patient counseling. In this study, OcularChat, an MLLM, was fine-tuned from Qwen2.5-VL using simulated patient-physician dialogues to diagnose age-related macular degeneration (AMD) through visual question answering on color fundus photographs (CFPs). A total of 705,850 simulated dialogues paired with 46,167 CFPs were generated to train OcularChat to identify key AMD features and produce reasoned predictions. OcularChat demonstrated strong classification performance in AREDS, achieving accuracies of 0.954, 0.849, and 0.678 for the three diagnostic tasks: advanced AMD, pigmentary abnormalities, and drusen size, significantly outperforming existing MLLMs. On AREDS2, OcularChat remained the top-performing method on all tasks. Across three independent ophthalmologist graders, OcularChat achieved higher mean scores than a strong baseline model for advanced AMD (3.503 vs. 2.833), pigmentary abnormalities (3.272 vs. 2.828), drusen size (3.064 vs. 2.433), and overall impression (2.978 vs. 2.464) on a 5-point clinical grading rubric. Beyond strong objective performance in AMD severity classification, OcularChat demonstrated the ability to provide diagnostic reasoning, clinically relevant explanations, and interactive dialogue, with high performance in subjective ophthalmologist evaluation. These findings suggest that MLLMs may enable accurate, interpretable, and clinically useful image-based diagnosis and classification of AMD. --- ### Cross-Lingual Jailbreak Detection via Semantic Codebooks Authors: Shirin Alanova, Bogdan Minko, Sabrina Sadiekh, Evgeniy Kokuykin Published: 2026-04-28 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.25716v1 Safety mechanisms for large language models (LLMs) remain predominantly English-centric, creating systematic vulnerabilities in multilingual deployment. Prior work shows that translating malicious prompts into other languages can substantially increase jailbreak success rates, exposing a structural cross-lingual security gap. We investigate whether such attacks can be mitigated through language-agnostic semantic similarity without retraining or language-specific adaptation. Our approach compares multilingual query embeddings against a fixed English codebook of jailbreak prompts, operating as a training-free external guardrail for black-box LLMs. We conduct a systematic evaluation across four languages, two translation pipelines, four safety benchmarks, three embedding models, and three target LLMs (Qwen, Llama, GPT-3.5). Our results reveal two distinct regimes of cross-lingual transfer. On curated benchmarks containing canonical jailbreak templates, semantic similarity generalizes reliably across languages, achieving near-perfect separability (AUC up to 0.99) and substantial reductions in absolute attack success rates under strict low-false-positive constraints. However, under distribution shift - on behaviorally diverse and heterogeneous unsafe benchmarks - separability degrades markedly (AUC $\approx$ 0.60-0.70), and recall in the security-critical low-FPR regime drops across all embedding models. --- ### Learning Generalizable Multimodal Representations for Software Vulnerability Detection Authors: Zeming Dong, Yuejun Guo, Qiang Hu, Yao Zhang, Maxime Cordy, Hao Liu, Mike Papadakis, Yongqiang Lyu Published: 2026-04-28 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2604.25711v2 Source code and its accompanying comments are complementary yet naturally aligned modalities-code encodes structural logic while comments capture developer intent. However, existing vulnerability detection methods mostly rely on single-modality code representations, overlooking the complementary semantic information embedded in comments and thus limiting their generalization across complex code structures and logical relationships. To address this, we propose MultiVul, a multimodal contrastive framework that aligns code and comment representations through dual similarity learning and consistency regularization, augmented with diverse code-text pairs to improve robustness. Experiments on widely adopted DiverseVul and Devign datasets across four large language models (LLMs) (i.e., DeepSeek-Coder-6.7B, Qwen2.5-Coder-7B, StarCoder2-7B, and CodeLlama-7B) show that MultiVul achieves up to 27.07% F1 improvement over prompting-based methods and 13.37% over code-only Fine-Tuning, while maintaining comparable inference efficiency. --- ### Adaptive Meta-Learning Stochastic Gradient Hamiltonian Monte Carlo Simulation for Bayesian Updating of Structural Dynamic Models Authors: Xianghao Meng, James L. Beck, Yong Huang, Hui Li Published: 2026-04-28 Categories: stat.AP, cs.LG, stat.ME, stat.ML arXiv: https://arxiv.org/abs/2604.25710v1 In the last few decades, Markov chain Monte Carlo (MCMC) methods have been widely applied to Bayesian updating of structural dynamic models in the field of structural health monitoring. Recently, several MCMC algorithms have been developed that incorporate neural networks to enhance their performance for specific Bayesian model updating problems. However, a common challenge with these approaches lies in the fact that the embedded neural networks often necessitate retraining when faced with new tasks, a process that is time-consuming and significantly undermines the competitiveness of these methods. This paper introduces a newly developed adaptive meta-learning stochastic gradient Hamiltonian Monte Carlo (AM-SGHMC) algorithm. The idea behind AM-SGHMC is to optimize the sampling strategy by training adaptive neural networks, and due to the adaptive design of the network inputs and outputs, the trained sampler can be directly applied to various Bayesian updating problems of the same type of structure without further training, thereby achieving meta-learning. Additionally, practical issues for the feasibility of the AM-SGHMC algorithm for structural dynamic model updating are addressed, and two examples involving Bayesian updating of multi-story building models with different model fidelity are used to demonstrate the effectiveness and generalization ability of the proposed method. --- ### Backtranslation Augmented Direct Preference Optimization for Neural Machine Translation Authors: Mehrdad Ghassabi, Spehr Rajabi, Hamidreza Baradaran Kashani, Sadra Hakim, Mahshid Keivandarian, Amirhossein Jahani Bahnamiri Published: 2026-04-28 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.25702v4 Contemporary neural machine translation (NMT) systems are almost exclusively built by training on supervised parallel data. Despite the tremendous progress achieved, these systems still exhibit persistent translation errors. This paper proposes that a post-training paradigm based on reinforcement learning (RL) can effectively rectify such mistakes. We introduce a novel framework that requires only a general text corpus and an expert translator which can be either human or an AI system to provide iterative feedback. In our experiments, we focus specifically on English-to-German translation as a representative high-resource language pair. Crucially, we implement this RL-based post-training using Direct Preference Optimization (DPO). Applying our DPO-driven framework to the gemma3-1b model yields a significant improvement in translation quality, elevating its COMET score from 0.703 to 0.747 on the English to German task. The results demonstrate that DPO offers an efficient and stable pathway for enhancing pre-trained NMT models through preference-based post-training. --- ### Bug-Report-Driven Fault Localization: Industrial Benchmarking and Lesson Learned at ABB Robotics Authors: Pernilla Hall, Anton Ununger, Riccardo Rubei, Alessio Bucaioni Published: 2026-04-28 Categories: cs.SE, cs.LG arXiv: https://arxiv.org/abs/2604.25700v2 Software quality assurance remains a major challenge in industrial environments, where large-scale and long-lived systems inevitably accumulate defects. Identifying the location of a fault is often time-consuming and costly, particularly during maintenance phases when developers must rely primarily on textual bug reports rather than complete runtime or code-level context. In this study, we investigated if artificial intelligence can support fault localization using only the natural-language content of bug reports. By relying only on textual information, our approach requires no access to source code, execution traces, or static analysis artifacts, making it directly deployable within existing industrial maintenance workflows. We framed fault localization as a supervised text classification problem and evaluated three traditional machine learning models (Logistic Regression, Support Vector Machine, and Random Forest) and two fine-tuned transformer-based language models (RoBERTa-Base and Distil-RoBERTa). Our evaluation used proprietary data from ABB Robotics in Sweden, comprising five years of resolved industrial bug reports, each linked to its verified code fix. This setting allowed us to assess model effectiveness under realistic industrial constraints. Our results showed that traditional models using term frequency-inverse document features consistently outperformed the fine-tuned language models on this dataset, while data augmentation improved Random Forest performance. These findings challenge the assumption that transformer-based models universally outperform classical approaches in industrial contexts with domain-specific data. We demonstrated that historical bug reports can be systematically used for text-based, artificial intelligence-assisted fault localization, providing a scalable, low-cost, and empirically grounded complement to common debugging practices in industry. --- ### RADD: Retrieval-Augmented Discrete Diffusion for Multi-Modal Knowledge Graph Completion Authors: Guanglin Niu, Bo Li Published: 2026-04-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2604.25693v2 Most multi-modal knowledge graph completion (MMKGC) models use one embedding scorer to conduct both retrieval over the full entity set and final link prediction. We argue that this coupling is a core bottleneck: global high-recall search and local fine-grained disambiguation require different inductive biases. Therefore, we propose a Retrieval-Augmented Discrete Diffusion (RADD) framework to decouple retrieval and reranking for MMKGC. A relation-aware multimodal knowledge graph embedding (KGE) retriever serves as both global retriever and distillation teacher, while a conditional discrete denoiser performs shortlist-level entity-identity generation for reranking. Training combines KGE supervision, denoising cross-entropy, and temperature-scaled distillation from the retriever to the denoiser. At inference, the designed Diff-Rerank first forms a top-K shortlist with the retriever and then reranks it with the denoiser, ensuring that recall is a strict requirement for precision. Experiments on three MMKGC benchmarks show that RADD achieves the best performance and consistent gains over strong unimodal, multimodal, and LLM-based baselines, while ablations further verify each component's contribution. --- ### Spreadsheet Modeling Experiments Using GPTs on Small Problem Statements and the Wall Task Authors: Thomas A. Grossman, Yuan Chen, Sopiko Datuashvili Published: 2026-04-28 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2604.25689v1 This paper investigates how GPT-based tools can assist in building reusable analytical spreadsheet models. After a screening, we evaluate five GPT extensions and select Excel AI by pulsrai.com for detailed testing. Through structured experiments on simple problem statements, we assess Excel AI's performance against the ERFR criteria (each input in a cell; cell formulas; no hardwired numbers; labels; accurate). Results show that while Excel AI can produce well-structured models, it is inconsistent and often non-reproducible. We identify two central challenges - "the problem of confidence" and "the problem of workflow" - which highlight the need for skilled users to verify and adapt GPT-generated spreadsheets. Though GPTs show promise for generating draft models that may reduce development time or lower skill requirements, current tools remain unreliable for professional use. We conclude with recommendations for future research into prompt engineering, reproducibility, and larger-scale modeling tasks. --- ### Learnable Burst Quantization for Expressive and Efficient Spiking Neural Networks Authors: Dewei Bai, Hongxiang Peng, Hong Qu, Dawen Xia Published: 2026-04-28 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.25688v2 Binary spikes provide only two neuronal output states per timestep, limiting the response capacity of spiking neural networks (SNNs) under short simulation horizons. Burst neurons expand this response space, but their threshold spacing is typically fixed before training, leaving layer-specific burst resolution outside end-to-end optimization. We propose Learnable Burst Quantization (LBQ), which formulates burst emission as saturated uniform quantization with a positive, layer-wise learnable step. ReLSG-ET, a rectified-linear surrogate gradient with exponential tails, provides gradient support throughout and beyond the active burst range, thereby enabling joint optimization of synaptic weights and burst resolution. At inference, LBQ absorbs each learned step into downstream weights and decomposes integer burst levels into binary bit planes, accumulating only the non-zero planes. This changes the synaptic accumulation count for a level $S$ from $S$ to $\operatorname{popcount}(S)$. At two timesteps, LBQ achieves 97.45\% on CIFAR-10 and 82.82\% on CIFAR-100 with ResNet-20, and 73.67\% on ImageNet-1K with ResNet-34. On CIFAR-10, it comes within 0.07 percentage points of the 97.52\% ResNet-20 ANN reference; at $N_{\max}=5$, bit-plane execution reduces unary-equivalent synaptic accumulations by 40.52\% relative to unary execution. Controlled ablations isolate the benefits of learned quantization and ReLSG-ET, while layer-wise analyses reveal selective burst allocation across network depth. Results on CIFAR10-DVS and DVS128-Gesture extend the evidence to event-driven recognition. LBQ therefore couples adaptive burst resolution and accurate inference with an algebraically equivalent bit-sparse synaptic execution path. --- ### QB-LIF: Learnable-Scale Quantized Burst Neurons for Efficient SNNs Authors: Dewei Bai, Hongxiang Peng, Jiajun Mei, Yang Ren, Hong Qu, Dawen Xia, Zhang Yi Published: 2026-04-28 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.25688v1 Binary spike coding enables sparse and event-driven computation in spiking neural networks (SNNs), yet its 1-bit-per-timestep representation fundamentally limits information throughput. This bottleneck becomes increasingly restrictive in deep architectures under short simulation horizons. We propose the Quantized Burst-LIF (QB-LIF) neuron, which reformulates burst spiking as a saturated uniform quantization of membrane potentials with a learnable scale. Instead of relying on predefined multi-threshold structures, QB-LIF treats the quantization scale as a trainable parameter, allowing each layer to autonomously adapt its spiking resolution to the underlying membrane-potential statistics. To preserve hardware efficiency, we introduce an absorbable scale strategy that folds the learned quantized scale into synaptic weights during inference, maintaining a strict accumulate-only (AC) execution paradigm. To enable stable optimization in the discrete multi-level space, we further design ReLSG-ET, a rectified-linear surrogate gradient with exponential tails that sustains gradient flow across burst intervals. Extensive experiments on static (CIFAR-10/100, ImageNet) and event-driven (CIFAR10-DVS, DVS128-Gesture) benchmarks demonstrate that QB-LIF consistently outperforms binary and fixed-burst SNNs, achieving higher accuracy under ultra-low latency while preserving neuromorphic compatibility. --- ### Robustness Evaluation of a Foundation Segmentation Model Under Simulated Domain Shifts in Abdominal CT: Implications for Health Digital Twin Deployment Authors: Sanghati Basu Published: 2026-04-28 Categories: eess.IV, cs.CV arXiv: https://arxiv.org/abs/2604.25685v1 Foundation segmentation models such as the Segment Anything Model (SAM) have demonstrated strong generalization across natural images; however, their robustness under clinically realistic medical imaging domain shifts remains insufficiently quantified. We present a systematic slice-level robustness audit of SAM (ViT-B) for spleen segmentation in abdominal CT using 1,051 nonempty slices from 41 volumes in the Medical Segmentation Decathlon. A standardized ground-truth-derived bounding-box protocol was used to isolate encoder robustness from prompt uncertainty. Controlled perturbations simulating inter-scanner variability, including Gaussian noise, blur, contrast scaling, gamma correction, and resolution mismatch, were applied across ten conditions. The clean baseline achieved a mean Dice score of 0.9145 (95% CI: (0.909, 0.919)) with a failure rate of 0.67%. Across all perturbations, the absolute mean ΔDice remained below 0.01. Paired Wilcoxon signed-rank tests with Benjamini-Hochberg false discovery rate correction identified statistically significant but small-magnitude changes under selected conditions, while McNemar analysis showed no significant increase in failure probability. These findings indicate that SAM exhibits stable segmentation behavior under moderate CT domain shifts, supporting its role as a robust foundation baseline for medical image segmentation research. As health digital twins increasingly incorporate foundation segmentation models for anatomical modeling and organ-level monitoring, formal characterization of robustness under real-world imaging variability is a necessary step toward trustworthy deployment. --- ### Think Before You Act -- A Neurocognitive Governance Model for Autonomous AI Agents Authors: Eranga Bandara, Ross Gore, Asanga Gunaratna, Sachini Rajapakse, Isurunima Kularathna, Ravi Mukkamala, Sachin Shetty, Xueping Liang, Amin Hass, Tharaka Hewa, Abdul Rahman, Christopher K. Rhea et al. Published: 2026-04-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2604.25684v1 The rapid deployment of autonomous AI agents across enterprise, healthcare, and safety-critical environments has created a fundamental governance gap. Existing approaches, runtime guardrails, training-time alignment, and post-hoc auditing treat governance as an external constraint rather than an internalized behavioral principle, leaving agents vulnerable to unsafe and irreversible actions. We address this gap by drawing on how humans self-govern naturally: before acting, humans engage deliberate cognitive processes grounded in executive function, inhibitory control, and internalized organizational rules to evaluate whether an intended action is permissible, requires modification, or demands escalation. This paper proposes a neurocognitive governance framework that formally maps this human self-governance process to LLM-driven agent reasoning, establishing a structural parallel between the human brain and the large language model as the cognitive core of an agent. We formalize a Pre-Action Governance Reasoning Loop (PAGRL) in which agents consult a four-layer governance rule set: global, workflow-specific, agent-specific, and situational before every consequential action, mirroring how human organizations structure compliance hierarchies across enterprise, department, and role levels. Implemented on a production-grade retail supply chain workflow, the framework achieves 95% compliance accuracy and zero false escalations to human oversight, demonstrating that embedding governance into agent reasoning produces more consistent, explainable, and auditable compliance than external enforcement. This work offers a principled foundation for autonomous AI agents that govern themselves the way humans do: not because rules are imposed upon them, but because deliberation is embedded in how they think. --- ### Exploring Remote Photoplethysmography for Neonatal Pain Detection from Facial Videos Authors: Ashutosh Dhamaniya, Anup Kumar Gupta, Trishna Saikia, Puneet Gupta Published: 2026-04-28 Categories: cs.CV, eess.IV arXiv: https://arxiv.org/abs/2604.25680v1 Unaddressed pain in neonates can lead to adverse effects, including delayed development and slower weight gain, emphasising the need for more objective and reliable pain assessment methods. Hence, automated methods using behavioural and physiological pain indicators have been developed to aid healthcare professionals in the Neonatal ICU. Traditional contact-based methods for physiological parameter estimation are unsuitable for long-term monitoring and increase the risk of spreading diseases like COVID-19. We introduce a novel approach using remote photoplethysmography (rPPG) to estimate pulse signals in a non-contact manner and employ them for neonatal pain detection. The temporal signals acquired from regions-of-interest (ROIs) affected by skin deformations may exhibit lower quality and provide erroneous rPPG signals. Therefore, we incorporated a quality parameter to select the temporal signals obtained from ROIs that are least affected by skin deformations. Further, we employed signal-to-noise ratio as a fitness parameter to extract the rPPG signal corresponding to the clip that is least affected by noise. Experimental findings demonstrate that the rPPG signals provide useful information for neonatal pain detection, and signals extracted from the blue colour channel outperform those extracted from other colour channels. We also show that combining rPPG and audio features provides better results than individual modalities. --- ### CORAL: Adaptive Retrieval Loop for Culturally-Aligned Multilingual RAG Authors: Nayeon Lee, Jiwoo Song, Byeongcheol Kang Published: 2026-04-28 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.25676v1 Multilingual retrieval-augmented generation (mRAG) is often implemented within a fixed retrieval space, typically via query or document translation or multilingual embedding vector representations. However, this approach may be inadequate for culturally grounded queries, in which retrieval-condition misalignment may occur. Even strong retrievers and generators may struggle to produce culturally relevant answers when sourcing evidence from inappropriate linguistic or regional contexts. To this end, we introduce CORAL (COntext-aware Retrieval with Agentic Loop, an adaptive retrieval methodology for mRAG that enables iterative refinement of both the retrieval space (corpora) and the retrieval probe (query) based on the quality of the evidence. The overall process includes: (1) selecting corpora, (2) retrieving documents, (3) critiquing evidence for relevance and cultural alignment, and (4) checking sufficiency. If the retrieved documents are insufficient to answer the query correctly, the system (5) reselects corpora and rewrites the query. Across two cultural QA benchmarks, CORAL achieves up to a 3.58%p accuracy improvement on low-resource languages relative to the strongest baselines. ---