# Research Papers — 2026-04 · bundle page 202 of 233 > Full-text markdown of papers #10051–10100 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/202 [← Previous bundle](https://aipod.dev/papers/archive/2026-04/page/201.md) [Next bundle →](https://aipod.dev/papers/archive/2026-04/page/203.md) ### Diagonal-Tiled Mixed-Precision Attention for Efficient Low-Bit MXFP Inference Authors: Yifu Ding, Xinhao Zhang, Jinyang Guo Published: 2026-04-05 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2604.03950v1 Transformer-based large language models (LLMs) have demonstrated remarkable performance across a wide range of real-world tasks, but their inference cost remains prohibitively high due to the quadratic complexity of attention and the memory bandwidth limitations of high-precision operations. In this work, we present a low-bit mixed-precision attention kernel using the microscaling floating-point (MXFP) data format, utilizing the computing capability on next-generation GPU architectures. Our Diagonal-Tiled Mixed-Precision Attention (DMA) incorporates two kinds of low-bit computation at the tiling-level, and is a delicate fused kernel implemented using Triton, exploiting hardware-level parallelism and memory efficiency to enable fast and efficient inference without compromising model performance. Extensive empirical evaluations on NVIDIA B200 GPUs show that our kernel maintains generation quality with negligible degradation, and meanwhile achieves significant speedup by kernel fusion. We release our code at https://github.com/yifu-ding/MP-Sparse-Attn. --- ### SafeCtrl: Region-Aware Safety Control for Text-to-Image Diffusion via Detect-Then-Suppress Authors: Lingyun Zhang, Yu Xie, Zhongli Fang, Yu Liu, Ping Chen Published: 2026-04-05 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.03941v1 The widespread deployment of text-to-image diffusion models is significantly challenged by the generation of visually harmful content, such as sexually explicit content, violence, and horror imagery. Common safety interventions, ranging from input filtering to model concept erasure, often suffer from two critical limitations: (1) a severe trade-off between safety and context preservation, where removing unsafe concepts degrades the fidelity of the safe content, and (2) vulnerability to adversarial attacks, where safety mechanisms are easily bypassed. To address these challenges, we propose SafeCtrl, a Region-Aware safety control framework operating on a Detect-Then-Suppress paradigm. Unlike global safety interventions, SafeCtrl first employs an attention-guided Detect module to precisely localize specific risk regions. Subsequently, a localized Suppress module, optimized via image-level Direct Preference Optimization (DPO), neutralizes harmful semantics only within the detected areas, effectively transforming unsafe objects into safe alternatives while leaving the surrounding context intact. Extensive experiments across multiple risk categories demonstrate that SafeCtrl achieves a superior trade-off between safety and fidelity compared to state-of-the-art methods. Crucially, our approach exhibits improved resilience against adversarial prompt attacks, offering a precise and robust solution for responsible generation. --- ### Fused Multinomial Logistic Regression Utilizing Summary-Level External Machine-learning Information Authors: Chi-Shian Dai, Jun Shao Published: 2026-04-05 Categories: stat.ME, cs.LG, stat.ML arXiv: https://arxiv.org/abs/2604.03939v1 In many modern applications, a carefully designed primary study provides individual-level data for interpretable modeling, while summary-level external information is available through black-box, efficient, and nonparametric machine-learning predictions. Although summary-level external information has been studied in the data integration literature, there is limited methodology for leveraging external nonparametric machine-learning predictions to improve statistical inference in the primary study. We propose a general empirical-likelihood framework that incorporates external predictions through moment constraints. An advantage of nonparametric machine-learning prediction is that it induces a rich class of valid moment restrictions that remain robust to covariate shift under a mild overlap condition without requiring explicit density-ratio modeling. We focus on multinomial logistic regression as the primary model and address common data-quality issues in external sources, including coarsened outcomes, partially observed covariates, covariate shift, and heterogeneity in generating mechanisms known as concept shift. We establish large-sample properties of the resulting fused estimator, including consistency and asymptotic normality under regularity conditions. Moreover, we provide mild sufficient conditions under which incorporating external predictions delivers a strict efficiency gain relative to the primary-only estimator. Simulation studies and an application to the National Health and Nutrition Examination Survey on multiclass blood-pressure classification. --- ### Biconvex Biclustering Authors: Sam Rosen, Eric C. Chi, Jason Xu Published: 2026-04-05 Categories: stat.ML, cs.LG, stat.ME arXiv: https://arxiv.org/abs/2604.03936v1 This article proposes a biconvex modification to convex biclustering in order to improve its performance in high-dimensional settings. In contrast to heuristics that discard a subset of noisy features a priori, our method jointly learns and accordingly weighs informative features while discovering biclusters. Moreover, the method is adaptive to the data, and is accompanied by an efficient algorithm based on proximal alternating minimization, complete with detailed guidance on hyperparameter tuning and efficient solutions to optimization subproblems. These contributions are theoretically grounded; we establish finite-sample bounds on the objective function under sub-Gaussian errors, and generalize these guarantees to cases where input affinities need not be uniform. Extensive simulation results reveal our method consistently recovers underlying biclusters while weighing and selecting features appropriately, outperforming peer methods. An application to a gene microarray dataset of lymphoma samples recovers biclusters matching an underlying classification, while giving additional interpretation to the mRNA samples via the column groupings and fitted weights. --- ### Supervised Dimensionality Reduction Revisited: Why LDA on Frozen CNN Features Deserves a Second Look Authors: Indar Kumar, Girish Karhana, Sai Krishna Jasti, Ankit Hemant Lade Published: 2026-04-05 Categories: cs.LG, cs.AI, cs.CV, stat.ML arXiv: https://arxiv.org/abs/2604.03928v2 Frozen pretrained image representations are widely used for transfer learning: a backbone is kept fixed, feature vectors are extracted, and a lightweight classifier is trained on top. This pipeline usually feeds the full feature vector to the classifier, even when the target task has far fewer classes than the pretraining task. We revisit a classical alternative: supervised dimensionality reduction with Linear Discriminant Analysis (LDA) before linear probing. We evaluate ten dimensionality-reduction strategies on frozen features from six backbones -- ResNet-18, ResNet-50, MobileNetV3-Small, EfficientNet-B0, ViT-B/16, and DINOv2-ViT-S/14 -- across CIFAR-100, Tiny ImageNet, and CUB-200-2011. Under a fixed logistic-regression protocol, LDA improves accuracy over full features in 11 of 12 coarse-grained configurations, with gains up to 4.5 percentage points while reducing feature dimensionality by 48-87%. The same projection consistently hurts on fine-grained CUB-200, where full features win across all six backbones. This establishes a practical boundary condition: LDA is useful when class-level structure is coarse enough to be captured by mean-separating directions, but it can discard subtle cues needed for fine-grained recognition. We also compare LDA with PCA, PCA+LDA, regularized LDA, Local Fisher Discriminant Analysis, Neighbourhood Components Analysis, and three lightweight LDA extensions. The results show that plain LDA offers the best accuracy-cost tradeoff for most coarse-grained settings, while more complex supervised reduction methods rarely justify their additional cost. Overall, the study provides concrete guidance for when post-hoc supervised projection should, and should not, be inserted into frozen-feature image classification pipelines. --- ### CODE-GEN: A Human-in-the-Loop RAG-Based Agentic AI System for Multiple-Choice Question Generation Authors: Xiaojing Duan, Frederick Nwanganga, Chaoli Wang Published: 2026-04-05 Categories: cs.AI, cs.CY, cs.HC, cs.MA arXiv: https://arxiv.org/abs/2604.03926v1 We present CODE-GEN, a human-in-the-Loop, retrieval-augmented generation (RAG)-based agentic AI system for generating context-aligned multiple-choice questions to develop student code reasoning and comprehension abilities. CODE-GEN employs an agentic AI architecture in which a Generator agent produces multiple-choice coding comprehension questions aligned with course-specific learning objectives, while a Validator agent independently assesses content quality across seven pedagogical dimensions. Both agents are augmented with specialized tools that enhance computational accuracy and verify code outputs. To evaluate the effectiveness of CODE-GEN, we conducted an evaluation study involving six human subject-matter experts (SMEs) who judged 288 AI-generated questions. The SMEs produced a total of 2,016 human-AI rating pairs, indicating agreement or disagreement with the assessments of Validator, along with 131 instances of qualitative feedback. Analyses of SME judgments show strong system performance, with human-validated success rates ranging from 79.9% to 98.6% across the seven pedagogical dimensions. The analysis of qualitative feedback reveals that CODE-GEN achieves high reliability on dimensions well suited to computational verification and explicit criteria matching, including question clarity, code validity, concept alignment, and correct answer validity. In contrast, human expertise remains essential for dimensions requiring deeper instructional judgment, such as designing pedagogically meaningful distractors and providing high-quality feedback that reinforces understanding. These findings inform the strategic allocation of human and AI effort in AI-assisted educational content generation. --- ### AdaptFuse: Training-Free Sequential Preference Learning via Externalized Bayesian Inference Authors: Fangzhou Lin, Peiran Li, Shuo Xing, Siyuan Yang, Qianwen Ge, Kazunori Yamada, Ziming Zhang, Haichong Zhang, Zhengzhong Tu Published: 2026-04-05 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.03925v1 Large language models struggle to accumulate evidence across multiple rounds of user interaction, failing to update their beliefs in a manner consistent with Bayesian inference. Existing solutions require fine-tuning on sensitive user interaction data, limiting their applicability in privacy-conscious settings. We propose AdaptFuse, a training-free framework that externalizes probabilistic computation entirely from the LLM: a symbolic module maintains a Bayesian posterior over a discrete hypothesis set, while a frozen LLM contributes semantic reasoning via multi-sample Dirichlet aggregation. The two signals are combined through entropy-adaptive fusion, which automatically weights each source by its predictive confidence, shifting reliance from the LLM to the symbolic posterior as evidence accumulates. We evaluate across three domains: flight recommendation, hotel recommendation, and web shopping; on Gemma 2 9B, Llama 3 8B, and Qwen 2.5 7B. AdaptFuse consistently outperforms both prompting baselines and fine-tuned Bayesian Teaching models on all tasks, with accuracy improving monotonically over interaction rounds. These results demonstrate that principled inference-time algorithms can substitute for fine-tuning in personalized recommendation, without storing or training on sensitive user data. All the code and materials will be open-sourced. --- ### Uncertainty as a Planning Signal: Multi-Turn Decision Making for Goal-Oriented Conversation Authors: Xinyi Ling, Ye Liu, Reza Averly, Xia Ning Published: 2026-04-05 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.03924v2 Goal-oriented conversational systems require making sequential decisions under uncertainty about the user's intent, where the algorithm must balance information acquisition and target commitment over multiple turns. Existing approaches address this challenge from different perspectives: structured methods enable multi-step planning but rely on predefined schemas, while LLM-based approaches support flexible interactions but lack long-horizon decision making, resulting in poor coordination between information acquisition and target commitment. To address this limitation, we formulate goal-oriented conversation as an uncertainty-aware sequential decision problem, where uncertainty serves as a guiding signal for multi-turn decision making. We propose a Conversation Uncertainty-aware Planning framework (CUP) that integrates language models with structured planning: a language model proposes feasible actions, and a planner evaluates their long-term impact on uncertainty reduction. Experiments on multiple conversational benchmarks show that CUP consistently improves success rates while requiring fewer interaction turns. Further analysis demonstrates that uncertainty-aware planning contributes to more efficient information acquisition and earlier confident commitment. --- ### Uncertainty as a Planning Signal: Multi-Turn Decision Making for Goal-Oriented Conversation Authors: Xinyi Ling, Ye Liu, Reza Averly, Xia Ning Published: 2026-04-05 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.03924v1 Goal-oriented conversational systems require making sequential decisions under uncertainty about the user's intent, where the algorithm must balance information acquisition and target commitment over multiple turns. Existing approaches address this challenge from different perspectives: structured methods enable multi-step planning but rely on predefined schemas, while LLM-based approaches support flexible interactions but lack long-horizon decision making, resulting in poor coordination between information acquisition and target commitment. To address this limitation, we formulate goal-oriented conversation as an uncertainty-aware sequential decision problem, where uncertainty serves as a guiding signal for multi-turn decision making. We propose a Conversation Uncertainty-aware Planning framework (CUP) that integrates language models with structured planning: a language model proposes feasible actions, and a planner evaluates their long-term impact on uncertainty reduction. Experiments on multiple conversational benchmarks show that CUP consistently improves success rates while requiring fewer interaction turns. Further analysis demonstrates that uncertainty-aware planning contributes to more efficient information acquisition and earlier confident commitment. --- ### ACES: Who Tests the Tests? Leave-One-Out AUC Consistency for Code Generation Authors: Hui Sun, Yun-Ji Zhang, Zheng Xie, Ren-Biao Liu, Yali Du, Xin-Ye Li, Ming Li Published: 2026-04-05 Categories: cs.LG arXiv: https://arxiv.org/abs/2604.03922v2 Selecting LLM-generated code candidates using LLM-generated tests is challenging because the tests themselves may be incorrect. Existing methods either treat all tests equally or rely on ad-hoc heuristics to filter unreliable tests. Yet determining test correctness requires knowing which codes are correct, creating a \emph{circular dependency}. Our key insight is that we need not determine test correctness at all: \emph{test votes should rank, not merely count}. What matters is not how many codes pass a test, but whether the test can \emph{distinguish} correct from incorrect code. We break the circular dependency via leave-one-out evaluation: hold out one test, rank codes by their aggregate scores on all remaining tests, and measure whether the held-out test's pass/fail pattern agrees with this ranking. We formalize this agreement as the leave-one-out AUC~(LOO-AUC) and prove that the expected LOO-AUC is proportional to each test's ability to separate correct code from incorrect code. Building on this, we propose \textbf{ACES}~(\textbf{A}UC \textbf{C}onsist\textbf{E}ncy \textbf{S}coring) with two complementary variants: ACES-C provides closed-form weights that provably approximate the oracle in expectation under a mild assumption on average test quality; ACES-O drops this assumption and iteratively optimizes a differentiable LOO-AUC objective. Both operate solely on the binary pass matrix with negligible overhead, and achieve state-of-the-art Pass@$k$ on multiple code generation benchmarks. --- ### From Plausible to Causal: Counterfactual Semantics for Policy Evaluation in Simulated Online Communities Authors: Agam Goyal, Yian Wang, Eshwar Chandrasekharan, Hari Sundaram Published: 2026-04-05 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.03920v2 LLM-based social simulations can generate believable community interactions, enabling ``policy wind tunnels'' where governance interventions are tested before deployment. But believability is not causality. Claims like ``intervention $A$ reduces escalation'' require causal semantics that current simulation work typically does not specify. We propose adopting the causal counterfactual framework, distinguishing \textit{necessary causation} (would the outcome have occurred without the intervention?) from \textit{sufficient causation} (does the intervention reliably produce the outcome?). This distinction maps onto different stakeholder needs: moderators diagnosing incidents require evidence about necessity, while platform designers choosing policies require evidence about sufficiency. We formalize this mapping, show how simulation design can support estimation under explicit assumptions, and argue that the resulting quantities should be interpreted as simulator-conditional causal estimates whose policy relevance depends on simulator fidelity. Establishing this framework now is essential: it helps define what adequate fidelity means and moves the field from simulations that look realistic toward simulations that can support policy changes. --- ### Interpreting Video Representations with Spatio-Temporal Sparse Autoencoders Authors: Atahan Dokme, Sriram Vishwanath Published: 2026-04-05 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2604.03919v2 We present the first systematic study of Sparse Autoencoders (SAEs) on video representations. Standard SAEs decompose video into interpretable, monosemantic features but destroy temporal coherence: hard TopK selection produces unstable feature assignments across frames, reducing autocorrelation by 36%. We propose spatio-temporal contrastive objectives and Matryoshka hierarchical grouping that recover and even exceed raw temporal coherence. The contrastive loss weight controls a tunable trade-off between reconstruction and temporal coherence. A systematic ablation on two backbones and two datasets shows that different configurations excel at different goals: reconstruction fidelity, temporal coherence, action discrimination, or interpretability. Contrastive SAE features improve action classification by +3.9 pp over raw features and text-video retrieval by up to 2.8 x R@1. A cross-backbone analysis reveals that standard monosemanticity metrics contain a backbone-alignment artifact: both DINOv2 and VideoMAE produce equally monosemantic features under an independent (CLIP) similarity space. Targeted feature ablation shows that contrastive training concentrates the probe's predictive signal into a small number of identifiable features. Supplementary material, code, configurations and evaluation scripts are available at https://github.com/atahandokme/spatio-temporal-sparse-autoencoders-video. --- ### Interpreting Video Representations with Spatio-Temporal Sparse Autoencoders Authors: Atahan Dokme, Sriram Vishwanath Published: 2026-04-05 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2604.03919v1 We present the first systematic study of Sparse Autoencoders (SAEs) on video representations. Standard SAEs decompose video into interpretable, monosemantic features but destroy temporal coherence: hard TopK selection produces unstable feature assignments across frames, reducing autocorrelation by 36%. We propose spatio-temporal contrastive objectives and Matryoshka hierarchical grouping that recover and even exceed raw temporal coherence. The contrastive loss weight controls a tunable trade-off between reconstruction and temporal coherence. A systematic ablation on two backbones and two datasets shows that different configurations excel at different goals: reconstruction fidelity, temporal coherence, action discrimination, or interpretability. Contrastive SAE features improve action classification by +3.9% over raw features and text-video retrieval by up to 2.8xR@1. A cross-backbone analysis reveals that standard monosemanticity metrics contain a backbone-alignment artifact: both DINOv2 and VideoMAE produce equally monosemantic features under neutral (CLIP) similarity. Causal ablation confirms that contrastive training concentrates predictive signal into a small number of identifiable features. --- ### Automating Cloud Security and Forensics Through a Secure-by-Design Generative AI Framework Authors: Dalal Alharthi, Ivan Roberto Kawaminami Garcia Published: 2026-04-05 Categories: cs.CR, cs.AI, cs.DC, cs.LG arXiv: https://arxiv.org/abs/2604.03912v1 As cloud environments become increasingly complex, cybersecurity and forensic investigations must evolve to meet emerging threats. Large Language Models (LLMs) have shown promise in automating log analysis and reasoning tasks, yet they remain vulnerable to prompt injection attacks and lack forensic rigor. To address these dual challenges, we propose a unified, secure-by-design GenAI framework that integrates PromptShield and the Cloud Investigation Automation Framework (CIAF). PromptShield proactively defends LLMs against adversarial prompts using ontology-driven validation that standardizes user inputs and mitigates manipulation. CIAF streamlines cloud forensic investigations through structured, ontology-based reasoning across all six phases of the forensic process. We evaluate our system on real-world datasets from AWS and Microsoft Azure, demonstrating substantial improvements in both LLM security and forensic accuracy. Experimental results show PromptShield boosts classification performance under attack conditions, achieving precision, recall, and F1 scores above 93%, while CIAF enhances ransomware detection accuracy in cloud logs using Likert-transformed performance features. Our integrated framework advances the automation, interpretability, and trustworthiness of cloud forensics and LLM-based systems, offering a scalable foundation for real-time, AI-driven incident response across diverse cloud infrastructures. --- ### Align Your Structures: Generating Trajectories with Structure Pretraining for Molecular Dynamics Authors: Aniketh Iyengar, Jiaqi Han, Pengwei Sun, Mingjian Jiang, Jianwen Xie, Stefano Ermon Published: 2026-04-05 Categories: cs.LG, q-bio.QM arXiv: https://arxiv.org/abs/2604.03911v1 Generating molecular dynamics (MD) trajectories using deep generative models has attracted increasing attention, yet remains inherently challenging due to the limited availability of MD data and the complexities involved in modeling high-dimensional MD distributions. To overcome these challenges, we propose a novel framework that leverages structure pretraining for MD trajectory generation. Specifically, we first train a diffusion-based structure generation model on a large-scale conformer dataset, on top of which we introduce an interpolator module trained on MD trajectory data, designed to enforce temporal consistency among generated structures. Our approach effectively harnesses abundant structural data to mitigate the scarcity of MD trajectory data and effectively decomposes the intricate MD modeling task into two manageable subproblems: structural generation and temporal alignment. We comprehensively evaluate our method on the QM9 and DRUGS small-molecule datasets across unconditional generation, forward simulation, and interpolation tasks, and further extend our framework and analysis to tetrapeptide and protein monomer systems. Experimental results confirm that our approach excels in generating chemically realistic MD trajectories, as evidenced by remarkable improvements of accuracy in geometric, dynamical, and energetic measurements. --- ### Improving Model Performance by Adapting the KGE Metric to Account for System Non-Stationarity Authors: M Jawad, HV Gupta, YH Wang, MA Farmani, A Behrangi, GY Niu Published: 2026-04-05 Categories: cs.LG, physics.ao-ph arXiv: https://arxiv.org/abs/2604.03906v2 Geoscientific systems tend to be characterized by pronounced temporal non-stationarity, arising from seasonal and climatic variability in hydrometeorological drivers, and from natural and anthropogenic changes to land use and cover. As has been pointed out, such variability renders "the assumption of statistical stationarity obsolete in water management", and requires us to "account for, rather than ignore, non-stationary trends" in the data. However, metrics used for model development are typically based on the implicit and unjustifiable assumption that the data generating process is time-stationary. Here, we introduce the JKGE_ss metric (adapted from KGE_ss) that detects and accounts for dynamical non-stationarity in the statistical properties of the data and thereby improves information extraction and model performance. Unlike NSE and KGE_ss, which use the long-term mean as a benchmark against which to evaluate model efficiency, JKGE_ss emphasizes reproduction of temporal variations in system storage. We tested the robustness of the new metric by training physical-conceptual and data-based catchment-scale models of varying complexity across a wide range of hydroclimatic conditions, from recent-precipitation-dominated to snow-dominated to strongly arid. In all cases, the result was improved reproduction of system temporal dynamics at all time scales, across wet to dry years, and over the full range of flow levels (especially recession periods). Since traditional metrics fail to adequately account for temporal shifts in system dynamics, potentially resulting in misleading assessments of model performance under changing conditions, we recommend the adoption of JKGE_ss for geoscientific model development. --- ### DC-Ada: Reward-Only Decentralized Sensor Adaptation for Heterogeneous Multi-Robot Teams Authors: Saad Alqithami Published: 2026-04-05 Categories: cs.RO, cs.AI, cs.MA arXiv: https://arxiv.org/abs/2604.03905v2 Heterogeneity is a defining feature of deployed multi-robot teams: platforms often differ in sensing modalities, ranges, fields of view, and failure patterns. Controllers trained under nominal sensing can degrade sharply when deployed on robots with missing or mismatched sensors, even when the task and action interface are unchanged. We present DC-Ada, a reward-only decentralized adaptation method that keeps a pretrained shared policy frozen and instead adapts compact per-robot observation transforms to map heterogeneous sensing into a fixed inference interface. DC-Ada is gradient-free and communication-minimal: it uses budgeted accept/reject random search with short common-random-number rollouts under a strict step budget. We evaluate DC-Ada against four baselines in a deterministic 2D multi-robot simulator covering warehouse logistics, search and rescue, and collaborative mapping, across four heterogeneity regimes (H0--H3) and five seeds with a matched budget of $200{,}000$ joint environment steps per run. Results show that heterogeneity can substantially degrade a frozen shared policy and that no single mitigation dominates across all tasks and metrics. Observation normalization is strongest for reward robustness in warehouse logistics and competitive in search and rescue, while the frozen shared policy is strongest for reward in collaborative mapping. DC-Ada offers a useful complementary operating point: it improves completion most clearly in severe coverage-based mapping while requiring only scalar team returns and no policy fine-tuning or persistent communication. These results position DC-Ada as a practical deploy-time adaptation method for heterogeneous teams. --- ### I-CALM: Incentivizing Confidence-Aware Abstention for LLM Selective Answering Authors: Haotian Zong, Binze Li, Yufei Long, Sinyin Chang, Jialong Wu, Gillian K. Hadfield Published: 2026-04-05 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.03904v2 Large language models (LLMs) often produce confident but incorrect answers, in part because standard evaluation incentives reward guessing over expressing uncertainty. We study epistemic abstention for factual questions with verifiable answers, where the goal is to improve selective answering, making LLMs abstain when they are likely to be wrong while preserving correct answers. Inspired by human behavioral decisions in question answering, we introduce I-CALM, a prompt-level framework for black-box LLMs. I-CALM combines elicited verbal confidence, announced answer/abstain payoffs, and normative guidance emphasizing truthfulness, humility, evidential support, and responsibility. To distinguish targeted abstention from indiscriminate refusal, we use a two-stage evaluation protocol, in which LLMs first choose whether to answer or abstain, and are then forced to provide a best guess for the abstained ones. Across models and factual QA datasets, I-CALM improves selective answering by reducing false-answer rate among surfaced responses and improving abstention quality, shifting error-prone cases into abstention while retaining answers the model would have answered correctly. Overall, I-CALM offers a lightweight way to improve inference-time selective answering without retraining or access to model's internal states. Code is available at https://github.com/FayLONG03/hallucinationControl. --- ### I-CALM: Incentivizing Confidence-Aware Abstention for LLM Hallucination Mitigation Authors: Haotian Zong, Binze Li, Yufei Long, Sinyin Chang, Jialong Wu, Gillian K. Hadfield Published: 2026-04-05 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.03904v1 Large language models (LLMs) frequently produce confident but incorrect answers, partly because common binary scoring conventions reward answering over honestly expressing uncertainty. We study whether prompt-only interventions -- explicitly announcing reward schemes for answer-versus-abstain decisions plus humility-oriented normative principles -- can reduce hallucination risk without modifying the model. Our focus is epistemic abstention on factual questions with a verifiable answer, where current LLMs often fail to abstain despite being uncertain about their answers. We first assess self-reported verbal confidence as a usable uncertainty signal, showing stability under prompt paraphrasing and reasonable calibration against a token-probability baseline. We then study I-CALM, a prompt-based framework that (i) elicits verbal confidence, (ii) partially rewards abstention through explicit reward schemes, and (iii) adds lightweight normative principles emphasizing truthfulness, humility, and responsibility. Using GPT-5 mini on PopQA as the main setting, we find that confidence-eliciting, abstention-rewarding prompts, especially with norms, reduce the false-answer rate on answered cases mainly by identifying and shifting error-prone cases to abstention and re-calibrating their confidence. This trades coverage for reliability while leaving forced-answer performance largely unchanged. Varying the abstention reward yields a clear abstention-hallucination frontier. Overall, results show the framework can improve selective answering on factual questions without retraining, with the magnitude of effect varying across models and datasets. Code is available at the following https://github.com/binzeli/hallucinationControl. --- ### Improving ML Attacks on LWE with Data Repetition and Stepwise Regression Authors: Alberto Alfarano, Eshika Saxena, Emily Wenger, François Charton, Kristin Lauter Published: 2026-04-05 Categories: cs.CR, cs.LG arXiv: https://arxiv.org/abs/2604.03903v1 The Learning with Errors (LWE) problem is a hard math problem in lattice-based cryptography. In the simplest case of binary secrets, it is the subset sum problem, with error. Effective ML attacks on LWE were demonstrated in the case of binary, ternary, and small secrets, succeeding on fairly sparse secrets. The ML attacks recover secrets with up to 3 active bits in the "cruel region" (Nolte et al., 2024) on samples pre-processed with BKZ. We show that using larger training sets and repeated examples enables recovery of denser secrets. Empirically, we observe a power-law relationship between model-based attempts to recover the secrets, dataset size, and repeated examples. We introduce a stepwise regression technique to recover the "cool bits" of the secret. --- ### Pseudo-Siamese Network for Planning in Target-Oriented Proactive Dialogues Authors: Xinyue Kang, Maodong Li, Yibin Zheng, Fang Kong Published: 2026-04-04 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2605.20195v1 A target-oriented proactive dialogue system is designed to steer conversations toward predefined targets while actively providing suggestions. The core paradigm of such a system is to plan a reasonable dialogue path and subsequently guide language models (e.g., pre-trained or large language models) to generate responses, where dialogue path planning serves as the central component-a novel yet under-explored problem. In this work, we propose a Forward-Focused Bidirectional Pseudo-Siamese Network (FF-BPSN) for dialogue path planning toward predefined dialogue targets. FF-BPSN employs two identical transformer-based decoders for forward and backward planning, together with a forward-focused module that integrates bidirectional information to construct the final forward path. This path benefits from bidirectional planning while prioritizing forward information. We then employ the planned path to guide language models in response generation. Extensive experiments on DuRecDial and DuRecDial 2.0 demonstrate that FF-BPSN achieves state-of-the-art performance in dialogue path planning and significantly enhances the effectiveness of target-oriented proactive dialogue systems. --- ### Parallel LLM Reasoning for Bias-Resilient, Robust Conceptual Abstraction Authors: Aisvarya Adeseye, Jouni Isoaho, Adeyemi Adeseye Published: 2026-04-04 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2605.20194v1 Large language models (LLMs) have been increasingly used to analyze text. However, they are often plagued with contextual reasoning limitations when analyzing long documents. When long documents are processed sequentially, early or dominant concepts can overshadow less visible but meaningful interpretations, leading to cumulative analytical bias, omission error, and over-generalization. Additionally, independently generated outputs are often merged without systematic grounding, introducing redundancy, conceptual drift, and unsupported claims. This study proposes a structured framework combining parallel chunk-level processing with evidence-anchored consolidation. Texts are first divided into semantically coherent chunks and processed independently in parallel to remove influence from earlier processing. The independently generated interpretations are then consolidated using explicit evidence anchoring and prioritization that reduces dominance and over-generalization while improving traceability. Experiments with multiple model types and sizes indicate that parallel processing significantly reduces omission error by approximately 84%, increases evidence traceability by up to 130%, and reduces unsupported claims by up to 91%. Smaller models benefited most, suggesting that efficient parallel chunking and consolidation play a critical role in achieving reliable and scalable textual analysis. --- ### Improving Quantized Model Performance in Qualitative Analysis with Multi-Pass Prompt Verification Authors: Aisvarya Adeseye, Jouni Isoaho, Adeyemi Adeseye Published: 2026-04-04 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2605.20193v1 Quantized Large Language Models (LLMs) are used more often in qualitative analysis because they run fast and need fewer computing resources. This study examines how different lower bits quantization levels (8-bit, 4-bit, 3-bit, and 2-bit) and quantization types affect the performance of LLaMA-3.1 (8B) on qualitative analysis. The study uses expert and non-expert responses from 82 interview transcripts. Low-bit models often produce higher levels of hallucinations and unstable results, especially when reading non-expert language with unclear terms. To improve performance, we propose a quantization-aware multi-pass prompt verification method. This method guides the model through controlled steps that reduce hallucinations. It removes unreliable content and passes the results to the next transcript after verification, improving accuracy. To validate performance, human coders analyzed transcripts using NVivo and BF16 LLaMA. BF16 LLaMA-3.1 produced high-precision output but had semantic drift and hallucination. These errors were corrected manually. The corrected BF16 output and NVivo human coding were combined to create a gold-standard ground truth (GSGT) for thematic extraction and frequency analysis. The results show that 8-bit models stay closest to the GSGT. The 4-bit models lose accuracy but become stable when the proposed method is applied. The 3-bit and 2-bit models drop in performance because of heavy compression, but they improve with the proposed prompt design and verification. The study also finds that models at the same bit level behave differently depending on quantization type. Overall, the method helps low-resource LLMs become more stable, accurate, and suitable for qualitative research at lower cost. --- ### Leveraging Large Language Models for Sentiment Analysis: Multi-Modal Analysis of Decentraland's MANA Token Authors: Xintong Wu, Peiting Tsai, Jing Yuan, Michael Yu, Greg Sun, Luyao Zhang Published: 2026-04-04 Categories: cs.CL, cs.CE, cs.CR, cs.CY, q-fin.CP arXiv: https://arxiv.org/abs/2605.20192v1 Decentraland, a decentralized virtual reality platform operating within the expanding Metaverse ecosystem, utilizes its native MANA token to facilitate virtual asset transactions and governance. This study investigates the integration of Discord community sentiment with multi-modal financial data to enhance cryptocurrency price prediction within virtual world economies. We address: (1) identifying sentiment patterns within Decentraland's Discord community, and (2) evaluating the impact of multi-modal features on token return forecasting. Using a BERT-based large language model for sentiment analysis, we develop two LSTM architectures: a baseline incorporating historical prices and a multi-modal variant integrating sentiment scores, trading volume, and market capitalization. Results indicate predominantly neutral community sentiment with a positive skew. The multi-modal model significantly outperforms the price-only baseline in prediction accuracy. These findings demonstrate the predictive value of community-derived signals for virtual economy forecasting and establish a foundation for future research at the intersection of immersive virtual environments, natural language processing, and cryptocurrency market analysis. --- ### Differences in Text Generated by Diffusion and Autoregressive Language Models Authors: Zeyang Zhang, Chengwei Liang, Xingyan Chen, Meiqi Gu, Minrui Luo, Jingzhao Zhang, Tianxing He Published: 2026-04-04 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2605.12522v1 Diffusion language models (DLMs) are promising alternatives to autoregressive language models (ARMs), yet the intrinsic differences in their generated text remain underexplored. We first find empirically that off-the-shelf DLMs exhibit lower $n$-gram entropy, higher semantic coherence, and higher semantic diversity. To understand the cause, we conduct controlled experiments that decouple the effects of training objectives and decoding algorithms. Results suggest that the DLM training objective contributes to the increases in semantic coherence and semantic diversity, but has a minor influence on entropy. These differences are primarily driven by the bidirectional context; other components in the training objective, such as input masking, label masking, and the weighting function, have a much weaker influence. Further, our experiments demonstrate that the reduction in entropy stems from DLMs' decoding algorithms, particularly confidence-based remasking strategies. We provide a theoretical understanding for this entropy reduction phenomenon. Together, our work uncovers key mechanisms underlying the differences between DLMs and ARMs in text generation, and informs future design of training objectives and decoding algorithms in DLMs. --- ### Adaptive Alarm Threshold Prediction in 4G Mobile Networks: A Percentile-Guided Deep Learning Framework with Interpretable Outputs Authors: Ayon Roy, Sadman Sharif, Shiva Prasad Sarkar Published: 2026-04-04 Categories: cs.NI, cs.LG arXiv: https://arxiv.org/abs/2605.00838v1 In mobile telecommunications, alarms act as early warning signals. They are triggered when a cell, the basic unit of radio coverage, shuts down or behaves abnormally. This signals a degradation in service quality, which directly affects the customer experience. To fix the issue, operators rely on preset thresholds to decide when an engineer should be sent out. In practice, these thresholds are set manually and remain fixed regardless of the time of day, traffic levels, or overall network conditions. This often leads to serious faults slipping through during busy hours, while minor issues can cause unnecessary callouts when the network is quiet. This paper presents a machine learning framework that automatically predicts four alarm thresholds, audit window duration, inactive time limit, total fluctuation count, and per hour fluctuation limit, from live network behavior. Since no ground truth labels exist for thresholds, we introduce a percentile guided label derivation strategy and evaluate four models on an anonymized dataset of 10,648 cells across three vendors and nine regions from a real 4G network, comprising a Gradient Boosted Trees baseline, a CNN-BiLSTM with attention, the proposed PCTN, and an iTransformer. PCTN performs the best overall with respect to three of the four targets, outperforming a state-of-the-art iTransformer while using 83 percent fewer parameters. Its mixed output heads and dynamic alpha mechanism produce thresholds that are both accurate and interpretable, allowing operators to inspect and adjust the learned policy without retraining. All comparisons are statistically significant at p < 0.001. The framework undergoes daily retraining using new data, which enables the thresholds to constantly adjust to changes in the network. --- ### Fast Log-Domain Sinkhorn Optimal Transport with Warp-Level GPU Reductions Authors: Hao Xiao Published: 2026-04-04 Categories: cs.LG arXiv: https://arxiv.org/abs/2605.00837v1 Entropic regularized optimal transport (OT) via the Sinkhorn algorithm has become a fundamental tool in machine learning, yet existing implementations either suffer from numerical instability for small regularization parameters or incur significant overhead from deep learning frameworks. We present FastSinkhorn, a lightweight, native CUDA implementation of the log-domain Sinkhorn algorithm that combines warp-level shuffle reductions with shared-memory tiling to achieve high GPU utilization without sacrificing numerical stability. Our solver operates entirely in the log-domain, enabling robust computation for regularization parameters as small as epsilon = 10^{-4} where standard-domain methods fail. On dense OT problems with n = m = 8192, our implementation achieves 12x speedup over the widely-used POT library and 5.9x speedup over GPU-accelerated PyTorch baselines, while consuming only 256 MB of GPU memory. We validate our solver on image color transfer, 3D point cloud matching, and convergence analysis, demonstrating that native CUDA kernels with careful numerical treatment provide a practical and efficient foundation for large-scale optimal transport computation. --- ### From Euler to Dormand-Prince: ODE Solvers for Flow Matching Generative Models Authors: Hao Xiao Published: 2026-04-04 Categories: cs.LG arXiv: https://arxiv.org/abs/2605.00836v1 Sampling from Flow Matching generative models requires solving an ordinary differential equation (ODE) whose computational cost is dominated by neural network forward passes. We derive four classical ODE solvers -- Euler, Explicit Midpoint, Classical Runge-Kutta (RK4), and Dormand-Prince 5(4) -- from first principles via Taylor expansion, implement them from scratch in PyTorch, and systematically benchmark their efficiency on Conditional Flow Matching tasks ranging from 2D toy distributions to MNIST digits. On the quantitative side, we use sliced Wasserstein distance to construct NFE-quality Pareto frontiers,finding that RK4 at 80 function evaluations achieves sample quality comparable to Euler at 200. Beyond reproducing known convergence rates, we report two empirical observations: (1) the Jacobian eigenvalue spectrum of the learned velocity field stiffens sharply near t=1, explaining why the adaptive Dormand-Prince solver automatically concentrates its step budget at the end of the trajectory; (2) the quality gap between low-order and high-order solvers widens for undertrained and smaller models, indicating that solver choice matters most when the model is imperfect. Code and all experiment scripts are publicly available. --- ### Sparse Regression under Correlation and Weak Signals: A Reproducible Benchmark of Classical and Bayesian Methods Authors: Hao Xiao Published: 2026-04-04 Categories: cs.LG arXiv: https://arxiv.org/abs/2605.00835v1 Choosing between classical and Bayesian sparse regression methods involves a real trade-off: penalized estimators like Lasso run in milliseconds but give no uncertainty estimates,while Horseshoe and Spike-and-Slab priors produce full posteriors but need MCMC chains that take minutes per fit.Surprisingly few studies compare these two families head-to-head under the conditions that actually make sparse regression hard -- correlated features, weak signals, and growing dimensionality. We benchmark six methods (OLS, Ridge,Lasso, Elastic Net, Horseshoe, Spike-and-Slab) on synthetic data with three covariance structures (rho up to 0.9), four SNR levels, and p in {20, 50, 100}, plus the Diabetes dataset,totalling over 2,600 experiments. The results are clear on some points and nuanced on others. Bayesian methods win on prediction error (MSE 72 vs. 108-267), and the Horseshoe delivers near-nominal 95% coverage (94.8%). But Spike-and-Slab,despite narrower intervals, under-covers at 91.9% -- its continuous relaxation likely plays a role. For variable selection, Lasso and Spike-and-Slab tie at F1 ~ 0.47, making Lasso the practical default when posteriors are not needed. Code and data are available at https://github.com/xiao98/sparse-bayesian-regression-bench. --- ### Polynomial-Time Optimal Group Selection via the Double-Commutator Eigenvalue Problem Authors: Mitchell A. Thornton Published: 2026-04-04 Categories: cs.LG, cs.CC, cs.IT arXiv: https://arxiv.org/abs/2605.00834v2 The algebraic diversity framework generalizes temporal averaging over multiple observations to algebraic group action on a single observation for second-order statistical estimation. The central open problem in this framework is $\textit{group selection}$: given an $M$-dimensional observation with unknown covariance structure, find the finite group whose spectral decomposition best matches the covariance. Naive enumeration of all subgroups of the symmetric group $S_M$ requires exponential time in $M$. We prove that this combinatorial problem reduces to a generalized eigenvalue problem derived from the double commutator of the covariance matrix, yielding a polynomial-time algorithm with complexity $O(d^2M^2 + d^3)$, where $d$ is the dimension of a generator basis. The minimum eigenvector of the double-commutator matrix directly constructs the optimal group generator in closed form, with no iterative optimization. The reduction is exact: the double-commutator minimum eigenvalue is zero if and only if the optimal generator lies in the span of the basis, and its magnitude provides a certifiable optimality gap when it does not. This problem does not appear in the standard catalogs of computational complexity (Garey and Johnson, 1979) and represents a new class linking group theory, matrix analysis, and statistical estimation. We establish connections to independent component analysis (JADE), structured matrix nearness problems, and simultaneous matrix diagonalization, and we show that the double-commutator formulation is the unique approach that is simultaneously polynomial-time, closed-form, and certifiable. We extend the framework to non-Abelian symmetry recovery via a Sequential GEVP with deflation, and add two identifiability theorems characterizing the commutant-lattice ambiguity and the dichotomy on whether $\mathrm{Aut}(\mathbf{R})$ recovers a generative subgroup or only a supergroup. --- ### GCA-BULF: A Bottom-Up Framework for Short-Term Load Forecasting Using Grouped Critical Appliances Authors: Yunhao Yao, Jinwei Fang, Puhan Luo, Zhiqiang Wang, Jiahui Hou, Xiang-Yang Li Published: 2026-04-04 Categories: cs.LG, cs.AI, eess.SP arXiv: https://arxiv.org/abs/2604.24766v1 With the rise of time-of-use and tiered electricity pricing, energy consumers are encouraged to adopt peak-shifting strategies by automatically controlling high-power appliances. These help lower energy costs while enhancing the power grid's stability. To support such energy management with high resilience and responsiveness, reliable short-term load forecasting (STLF) plays a critical role. STLF predicts electricity consumption over time horizons ranging from minutes to days, using historical data, temporal patterns, and contextual factors. Traditional top-down forecasting methods struggle to capture the complex consumption patterns of diverse and mixed appliance loads. Although bottom-up methods improve forecasting accuracy by integrating appliance-level data, monitoring all appliances is costly, and many do not meaningfully impact total load prediction. Therefore, we propose GCA-BULF, a bottom-up short-term load forecasting framework based on grouped critical appliances, supported by three key designs. First, the Critical Appliance Filtering module ranks appliances according to their power consumption, switching frequency, and usage pattern periodicity, and identifies critical ones through iterative load decomposition. Next, the Related Appliance Grouping module clusters these appliances based on spatial and temporal correlations for group-level forecasting. Finally, the Collaborative Load Forecasting module refines the total load prediction by combining multiple group-level forecasts. We evaluate GCA-BULF on residential and office building load forecasting tasks. Experimental results reveal that GCA-BULF improves hourly total load forecasting by 20.85%-57.88% compared to existing top-down methods and by 33.03%-92.48% compared to bottom-up methods. --- ### Interpretable Fuzzy Modeling Reveals Population-Level Representation Differences in P300 Brain Computer Interfaces Across Neurodivergent and Neurotypical Cohorts Authors: Xiaowei Jiang, Sudong Shang, Adrian Wilkinson, Michael L. Platt, Da Xiao, Bening Cao, Thomas Do Published: 2026-04-04 Categories: eess.SP, cs.HC, cs.LG arXiv: https://arxiv.org/abs/2604.24765v1 P300-based brain-computer interfaces (BCIs) are widely used for communication, but population heterogeneity may alter the neural patterns available for decoding. Prior work has mainly examined such differences at the signal or performance level, while the representation structure learned by the decoder remains underexplored. In this study, we propose an interpretable fuzzy spatiotemporal framework for P300 classification and use it to analyze population-level differences across amyotrophic lateral sclerosis (ALS), autism (AUT), and neurotypical (NT) cohorts. The model employs spatial and temporal fuzzy filters with learnable prototypes, enabling both classification and reconstruction of cohort-specific fuzzy centers. Experiments were conducted on ALS and NT subsets from bigP3BCI and on the BCIAUT-P300 benchmark in a within-subject setting. The proposed model achieved competitive performance against multiple deep learning baselines. More importantly, the reconstructed fuzzy centers revealed systematic cohort-dependent differences in waveform morphology and representation geometry. Point-wise statistical analysis identified significant temporal differences between cohorts, including intervals overlapping with the canonical P300 window, and low-dimensional embeddings showed partially separated cohort-specific prototype organizations. These results suggest that population heterogeneity in P300-BCI is reflected not only in decoding performance but also in the discriminative structure learned by the model. The proposed framework provides an interpretable route toward population-aware P300-BCI analysis and design. --- ### AutoCompress: Critical Layer Isolation for Efficient Transformer Compression Authors: Archit Thorat Published: 2026-04-04 Categories: cs.LG arXiv: https://arxiv.org/abs/2604.22786v1 We present AutoCompress, a transformer compression method motivated by an empirical finding: in small transformers, Layer 0 carries disproportionately high task-critical information, with an NTK-based importance score of 3.6 compared to a maximum of 0.054 for all other layers -- a gap of over 60x. Based on this finding, we propose Critical Layer Isolation (CLI), an architecture that protects Layer 0 at full dimensionality, compresses all intermediate layers through a learned bottleneck, and restores the full dimension at the final layer. Applied to GPT-2 Medium (354.8M parameters), CLI-GPT2 achieves 204.5 perplexity on WikiText-103 with only 143.8M parameters -- a 2.47x compression ratio and 59.5% parameter reduction. Crucially, an ablation study demonstrates that a uniform bottleneck baseline of comparable size achieves only 571.8 perplexity under identical training conditions, confirming that the architectural decision to protect Layer 0 -- rather than simply reducing model size -- is the primary driver of performance. Code and checkpoints are publicly available. --- ### Functional Similarity Metric for Neural Networks: Overcoming Parametric Ambiguity via Activation Region Analysis Authors: Kutomanov Hennadii Published: 2026-04-04 Categories: cs.LG arXiv: https://arxiv.org/abs/2604.16426v1 As modern deep learning architectures grow in complexity, representational ambiguity emerges as a critical barrier to their interpretability and reliable merging. For ReLU networks, identical functional mappings can be achieved through entirely different weight configurations due to algebraic symmetries: neuron permutation and positive diagonal scaling. Consequently, traditional parameter-based comparison methods exhibit extreme instability to slight weight perturbations during training. This paper proposes a mathematically grounded approach to constructing a stable canonical representation of neural networks and a robust functional similarity metric. We shift focus from comparing raw weights to analyzing the topology of neuron activation regions. The algorithm first eliminates scaling ambiguity via L2-normalization of weight vectors with subsequent layer compensation. Next, discrete approximations of activation regions are generated as binary functional signatures evaluated over a data sample. To overcome the computational bottleneck of comparing large binary vectors, we adapt Locality-Sensitive Hashing, specifically MinHash, providing a fast and statistically precise approximation of the Jaccard index. The final cross-network neuron matching is formulated as a linear sum assignment problem solved via the Hungarian algorithm. We demonstrate theoretically and experimentally that our metric mitigates the neuron "flickering" effect and exhibits exceptional robustness to minor weight perturbations. This framework provides a solid foundation for model merging, transfer learning, objective assessment during pruning, and Explainable AI paradigms. --- ### Method for Aggregating Unstructured Data Using Large Language Models Authors: Vsevolod Lazebnyi, Natalia Tereshkina, Maria Shabarina, Dmitriy Fedorov Published: 2026-04-04 Categories: cs.DB, cs.LG arXiv: https://arxiv.org/abs/2604.16425v1 This paper presents a method for the automated collection and aggregation of unstructured data from diverse web sources, utilizing Large Language Models (LLMs). The primary challenge with existing techniques is their instability when the structure of webpages changes, their limited support for dynamically loaded content during information collection, and the requirement for labor-intensive manual design of data pre-processing processes. The proposed algorithm integrates hybrid web scraping (Goose3 for static pages and Selenium+WebDriver for dynamic ones), data storage in a non-relational MongoDB database management system (DBMS), and intelligent extraction and normalization of information using LLMs into a predetermined JSON schema. A key scientific contribution of this study is a two-stage verification process for the generated data, designed to eliminate potential hallucinations byy comparing the embeddings of multiple LLM outputs obtained with different temperature parameter values, combined with formalized rules for monitoring data consistency and integrity. The experimental findings indicate a high level of accuracy in the completion of key fields, as well as the robustness of the proposed methodology to changes in web page structures. This makes it suitable for use in tasks such as news content aggregation, monitoring, and log analysis in near real-time mode, with the capacity to scale rapidly in terms of the number of sources. --- ### Safety, Security, and Cognitive Risks in State-Space Models: A Systematic Threat Analysis with Spectral, Stateful, and Capacity Attacks Authors: Manoj Parmar Published: 2026-04-04 Categories: cs.CR, cs.AI, cs.CL, cs.LG, math.OC arXiv: https://arxiv.org/abs/2604.16424v1 State-Space Models (SSMs) -- structured SSMs (S4, S4D, DSS, S5), selective SSMs (Mamba, Mamba-2), and hybrid architectures (Jamba) -- are deployed in safety-critical long-context applications: genomic analysis, clinical time-series forecasting, and cybersecurity log processing. Their linear-time scaling is compelling, yet the security properties of their compressed-state recurrent architectures remain unstudied. We present the first systematic treatment of SSM safety, security, and cognitive risks. Seven contributions: (1) Formal threat framework -- SSM Attack Surface (five layers), State Integrity Violation (StIV), Cross-Context Amplification Ratio $\mathcal{X}_\mathcal{S}$, and a Spectral Sensitivity Proposition grounded in the $H_\infty$ norm. (2) Three novel attack classes: spectral adversarial attacks (transfer-function gain exploitation), delayed-trigger stateful backdoors (activate thousands of steps after injection), and state capacity saturation (entropy flooding forces silent forgetting). (3) 14 MITRE ATLAS technique extensions across the full tactic chain. (4) Six-profile attacker taxonomy with kill chains for genomics, clinical, and cybersecurity domains. (5) Four cognitive risk hypotheses grounded in state-compression mechanics. (6) Governance-aligned mitigations mapped to CREST, NIST AI 600-1, and EU AI Act. (7) Empirical evaluation: targeted genomic injection achieves $\mathrm{StIV}=0.519$ vs. $0.086$ random ($6.0\times$, $p<0.001$); PGD state injection achieves $156\times$ output perturbation over random; SSD-structured extraction confirmed at $O(N^2)$ vs. $O(N^3)$ query complexity ($N\times$ speedup). Validation on pretrained checkpoints is detailed in the Appendix. --- ### Aletheia: Gradient-Guided Layer Selection for Efficient LoRA Fine-Tuning Across Architectures Authors: Abdulmalek Saket Published: 2026-04-04 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2604.15351v1 Low-Rank Adaptation (LoRA) has become the dominant parameter-efficient fine-tuning method for large language models, yet standard practice applies LoRA adapters uniformly to all transformer layers regardless of their relevance to the downstream task. We introduce Aletheia, a gradient-guided layer selection method that identifies the most task-relevant layers via a lightweight gradient probe and applies LoRA adapters only to those layers with asymmetric rank allocation. Across 81 experiment rows covering 14 successful models from 8 architecture families (0.5B-72B parameters, including dense and Mixture-of-Experts architectures), with one additional documented failed Pythia/GPT-NeoX attempt in Campaign 2, Aletheia achieves a 15-28% training speedup (mean 23.1%, p < 0.001) with bounded extra forgetting and broadly matched downstream behavior on the evaluated MMLU, GSM8K, and HumanEval benchmark pack. Across the tested families and scales, Campaign 1 shows a 100% per-model speed win rate and Campaign 2 shows broadly preserved downstream behavior within a bounded-degradation framing. Together these results support a practical model-economics claim: intelligent layer selection can make LoRA fine-tuning materially more efficient without introducing major downstream damage on the evaluated set. --- ### ML-based approach to classification and generation of structured light propagation in turbulent media Authors: Aokun Wang, Anjali Nair, Zhongjian Wang, Guillaume Bal Published: 2026-04-04 Categories: physics.optics, cs.LG, math.OC, physics.comp-ph arXiv: https://arxiv.org/abs/2604.14208v2 We study the classification task of structured-light beams after propagation through a random turbulent medium. The received speckle patterns are generated by numerical simulation of a stochastic paraxial propagation model, and the classification task is formulated over a finite alphabet of 15 OAM source classes. We benchmark intensity and autocorrelation inputs using SimpleCNN and ResNet-18 as classifiers. We also quantify the effect of training-set size and receiver-window misalignment. Since additional propagated samples may be costly to obtain, we develop a class-conditioned diffusion model for generative augmentation of turbulence-degraded intensity images. The main contribution is a spectrum-aware diffusion objective: a pixel-domain loss combined with a Fourier-domain Bregman regularizer designed to preserve high-frequency speckle statistics. We prove that this hybrid objective is consistent with the posterior-mean regression target of the diffusion model and show that generated samples substantially improve low-data classification. --- ### Portfolio Optimization Proxies under Label Scarcity and Regime Shifts via Bayesian and Deterministic Students under Semi-Supervised Sandwich Training Authors: Adhiraj Chattopadhyay Published: 2026-04-04 Categories: cs.LG, q-fin.PM, stat.ML arXiv: https://arxiv.org/abs/2604.14206v1 This paper proposes a machine learning assisted portfolio optimization framework designed for low data environments and regime uncertainty. We construct a teacher student learning pipeline in which a Conditional Value at Risk (CVaR) optimizer generates supervisory labels, and neural models (Bayesian and deterministic) are trained using both real and synthetically augmented data. The synthetic data is generated using a factor based model with t copula residuals, enabling training beyond the limited real sample of 104 labeled observations. We evaluate four student models under a structured experimental framework comprising (i) controlled synthetic experiments (3 x 5 seed grid), (ii) in-distribution real market evaluation (C2A) and (iii) cross-universe generalization (D2A). In real-market settings, models are deployed using a rolling evaluation protocol where a frozen pretrained model is periodically fine tuned on recent observations and reset to its base state, ensuring stability while allowing limited adaptation. Results show that student models can match or outperform the CVaR teacher in several settings, while achieving improved robustness under regime shifts and reduced turnover. These findings suggest that hybrid optimization learning approaches can enhance portfolio construction in data constrained environments --- ### Design Conditions for Intra-Group Learning of Sequence-Level Rewards: Token Gradient Cancellation Authors: Fei Ding, Yongkang Zhang, youwei wang, Zijian Zeng Published: 2026-04-04 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2604.13088v2 Reinforcement learning for multi-step reasoning with large language models (LLMs) typically relies on sparse terminal rewards, which creates a poorly conditioned credit-assignment problem: the final feedback is propagated uniformly across all intermediate decisions. This leads to high gradient variance, unstable training, and many ineffective updates, ultimately limiting sustained model improvement. We propose a counterfactual-comparison framework for credit assignment. For each input, the framework samples multiple reasoning trajectories and treats their differences as implicit approximations to alternative decisions. This yields an implicit process-level advantage estimator that converts sparse terminal rewards into step-sensitive learning signals. Building on this framework, we introduce Implicit Behavior Policy Optimization (IBPO), which substantially improves training stability and the performance ceiling on mathematical and code-reasoning benchmarks. Our results point to a promising direction for unlocking the reasoning potential of LLMs. --- ### Adaptive Rigor in AI System Evaluation using Temperature-Controlled Verdict Aggregation via Generalized Power Mean Authors: Aleksandr Meshkov Published: 2026-04-04 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.08595v1 Existing evaluation methods for LLM-based AI systems, such as LLM-as-a-Judge, verdict systems, and NLI, do not always align well with human assessment because they cannot adapt their strictness to the application domain. This paper presents Temperature-Controlled Verdict Aggregation (TCVA), a method that combines a five-level verdict-scoring system with generalized power-mean aggregation and an intuitive temperature parameter T (0.1, 1.0) to control evaluation rigor. Low temperatures yield pessimistic scores suited for safety-critical domains; high temperatures produce lenient scores appropriate for conversational AI. Experimental evaluation on three benchmark datasets with human Likert-scale annotations (SummEval and USR) shows that TCVA achieves correlation with human judgments comparable to RAGAS on faithfulness (Spearman = 0.667 vs. 0.676) while consistently outperforming DeepEval. The method requires no additional LLM calls when adjusting the temperature parameter. --- ### Negotiating Privacy with Smart Voice Assistants: Risk-Benefit and Control-Acceptance Tensions Authors: Molly Campbell, Mohamad Sheikho Al Jasem, Ajay Kumar Shrestha Published: 2026-04-04 Categories: cs.CR, cs.AI, cs.CY arXiv: https://arxiv.org/abs/2604.06235v1 Smart Voice assistants (SVAs) are widely adopted by youth, yet privacy decision-making in these environments is often characterized by competing considerations rather than clear-cut preferences. While our prior research has examined privacy risks, benefits, trust, and self-efficacy as distinct predictors of behavior, less attention has been paid to how these factors combine into higher-level tension that shapes privacy outcomes. This study introduces a negotiation-based framework for understanding youth privacy decision-making with SVAs by operationalizing two composite indices: the Risk-Benefit Tension Index (RBTI) and the Control-Acceptance Tension Index (CATI), using survey data from 469 Canadian youth aged 16-24. We examine the distribution of these indices and their relationship with privacy-protective behavior and SVA usage. Results show that both indices are meaningfully associated with protective action. Frequent SVA usage exhibits more benefit-dominant and acceptance-leaning negotiation profiles, suggesting that convenience-driven engagement may come at the expense of perceived control. By reframing privacy decision-making as a process of negotiation rather than inconsistency, this study offers a complementary perspective on the privacy paradox and provides a compact measurement approach for capturing how youth navigate competing privacy pressures in voice-enabled ecosystems. --- ### EDGE-Shield: Efficient Denoising-staGE Shield for Violative Content Filtering via Scalable Reference-Based Matching Authors: Takara Taniguchi, Ryohei Shimizu, Duc Minh Vo, Kota Izumi, Shiqi Yang, Teppei Suzuki Published: 2026-04-04 Categories: cs.CV, cs.MM arXiv: https://arxiv.org/abs/2604.06063v2 The advent of Text-to-Image generative models poses significant risks of copyright violation and deepfake generation. Since the rapid proliferation of new copyrighted works and private individuals constantly emerges, reference-based training-free content filters are essential for providing up-to-date protection without the constraints of a fixed knowledge cutoff. However, existing reference-based approaches often lack scalability when handling numerous references and require waiting for finishing image generation. To solve these problems, we propose EDGE-Shield, a scalable content filter during the denoising process that maintains practical latency while effectively blocking violative content. We leverage embedding-based matching for efficient reference comparison. Additionally, we introduce an \textit{$x$}-pred transformation that converts the model's noisy intermediate latent into the pseudo-estimated clean latent at the later stage, enhancing classification accuracy of violative content at earlier denoising stages. We conduct experiments of violative content filtering against two generative models including Z-Image-Turbo and Qwen-Image. EDGE-Shield significantly outperforms traditional reference-based methods in terms of latency; it achieves an approximate $79\%$ reduction in processing time for Z-Image-Turbo and approximate $50\%$ reduction for Qwen-Image, maintaining the filtering accuracy across different model architectures. --- ### Territory Paint Wars: Diagnosing and Mitigating Failure Modes in Competitive Multi-Agent PPO Authors: Diyansha Singh Published: 2026-04-04 Categories: cs.LG arXiv: https://arxiv.org/abs/2604.04983v1 We present Territory Paint Wars, a minimal competitive multi-agent reinforcement learning environment implemented in Unity, and use it to systematically investigate failure modes of Proximal Policy Optimisation (PPO) under self-play. A first agent trained for $84{,}000$ episodes achieves only $26.8\%$ win rate against a uniformly-random opponent in a symmetric zero-sum game. Through controlled ablations we identify five implementation-level failure modes -- reward-scale imbalance, missing terminal signal, ineffective long-horizon credit assignment, unnormalised observations, and incorrect win detection -- each of which contributes critically to this failure in this setting. After correcting these issues, we uncover a distinct emergent pathology: competitive overfitting, where co-adapting agents maintain stable self-play performance while generalisation win rate collapses from $73.5\%$ to $21.6\%$. Critically, this failure is undetectable via standard self-play metrics: both agents co-adapt equally, so the self-play win rate remains near $50\%$ throughout the collapse. We propose a minimal intervention -- opponent mixing, where $20\%$ of training episodes substitute a fixed uniformly-random policy for the co-adaptive opponent -- which mitigates competitive overfitting and restores generalisation to $77.1\%$ ($\pm 12.6\%$, $10$ seeds) without population-based training or additional infrastructure. We open-source Territory Paint Wars to provide a reproducible benchmark for studying competitive MARL failure modes. --- ### CURE:Circuit-Aware Unlearning for LLM-based Recommendation Authors: Ziheng Chen, Jiali Cheng, Zezhong Fan, Hadi Amiri, Yunzhi Yao, Xiangguo Sun, Yang Zhang Published: 2026-04-04 Categories: cs.IR, cs.AI, cs.CL, cs.LG arXiv: https://arxiv.org/abs/2604.04982v1 Recent advances in large language models (LLMs) have opened new opportunities for recommender systems by enabling rich semantic understanding and reasoning about user interests and item attributes. However, as privacy regulations tighten, incorporating user data into LLM-based recommendation (LLMRec) introduces significant privacy risks, making unlearning algorithms increasingly crucial for practical deployment. Despite growing interest in LLMRec unlearning, most existing approaches formulate unlearning as a weighted combination of forgetting and retaining objectives while updating model parameters in a uniform manner. Such formulations inevitably induce gradient conflicts between the two objectives, leading to unstable optimization and resulting in either ineffective unlearning or severe degradation of model utility. Moreover, the unlearning procedure remains largely black-box, undermining its transparency and trustworthiness. To tackle these challenges, we propose CURE, a circuit-aware unlearning framework that disentangles model components into functionally distinct subsets and selectively updates them. Here, a circuit refers to a computational subgraph that is causally responsible for task-specific behaviors. Specifically, we extract the core circuits underlying item recommendation and analyze how individual modules within these circuits contribute to the forget and retain objectives. Based on this analysis, these modules are categorized into forget-specific, retain-specific, and task-shared groups, each subject to function-specific update rules to mitigate gradient conflicts during unlearning. Experiments on real-world datasets show that our approach achieves more effective unlearning than existing baselines. --- ### An Imbalanced Dataset with Multiple Feature Representations for Studying Quality Control of Next-Generation Sequencing Authors: Philipp Röchner, Clarissa Krämer, Johannes U Mayer, Franz Rothlauf, Steffen Albrecht, Maximilian Sprang Published: 2026-04-04 Categories: q-bio.GN, cs.LG, cs.NE arXiv: https://arxiv.org/abs/2604.04981v2 Next-generation sequencing (NGS) is a key technique for studying the DNA and RNA of organisms. However, identifying quality problems in NGS data across different experimental settings remains challenging. To develop automated quality-control tools, researchers require datasets with features that capture the characteristics of quality problems. Existing NGS repositories, however, offer only a limited number of quality-related features. To address this gap, we propose a dataset derived from 37,491 NGS samples with two types of quality-related feature representations. The first type consists of 34 features derived from quality control tools (QC-34 features). The second type has a variable number of features ranging from eight to 1,183. These features were derived from read counts in problematic genomic regions identified by the ENCODE blocklist (BL features). All features describe the same human and mouse samples from five genomic assays, allowing direct comparison of feature representations. The proposed dataset includes a binary quality label, derived from automated quality control and domain experts. Among all samples, $3.2\%$ are of low quality. Supervised machine learning algorithms accurately predicted quality labels from the features, confirming the relevance of the provided feature representations. The proposed feature representations enable researchers to study how different feature types (QC-34 vs. BL features) and granularities (varying number of BL features) affect the detection of quality problems. --- ### Squeez: Task-Conditioned Tool-Output Pruning for Coding Agents Authors: Ádám Kovács Published: 2026-04-04 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2604.04979v1 Coding agents repeatedly consume long tool observations even though only a small fraction of each observation matters for the next step. We study task-conditioned tool-output pruning: given a focused query and one tool output, return the smallest verbatim evidence block the agent should inspect next. We introduce a benchmark of 11,477 examples built from SWE-bench repository interactions and synthetic multi-ecosystem tool outputs, with a manually curated 618-example test set. We fine-tune Qwen 3.5 2B with LoRA and compare it against larger zero-shot models and heuristic pruning baselines. Our model reaches 0.86 recall and 0.80 F1 while removing 92% of input tokens, outperforming zero-shot Qwen 3.5 35B A3B by 11 recall points and all heuristic baselines by a wide margin. --- ### Measuring the Permission Gate: A Stress-Test Evaluation of Claude Code's Auto Mode Authors: Zimo Ji, Zongjie Li, Wenyuan Jiang, Yudong Gao, Shuai Wang Published: 2026-04-04 Categories: cs.SE, cs.AI, cs.CR arXiv: https://arxiv.org/abs/2604.04978v2 Claude Code's auto mode is the first deployed permission system for AI coding agents, using a two-stage transcript classifier to gate dangerous tool calls. Anthropic reports a 0.4% false positive rate and 17% false negative rate on production traffic. We present the first independent evaluation of this system on deliberately ambiguous authorization scenarios, i.e., tasks where the user's intent is clear but the target scope, blast radius, or risk level is underspecified. Using AmPermBench, a 128-prompt benchmark spanning four DevOps task families and three controlled ambiguity axes, we evaluate 253 state-changing actions at the individual action level against oracle ground truth. Our findings characterize auto mode's scope-escalation coverage under this stress-test workload. The end-to-end false negative rate is 81.0% (95% CI: 73.8%-87.4%), substantially higher than the 17% reported on production traffic, reflecting a fundamentally different workload rather than a contradiction. Notably, 36.8% of all state-changing actions fall outside the classifier's scope via Tier 2 (in-project file edits), contributing to the elevated end-to-end FNR. Even restricting to the 160 actions the classifier actually evaluates (Tier 3), the FNR remains 70.3%, while the FPR rises to 31.9%. The Tier 2 coverage gap is most pronounced on artifact cleanup (92.9% FNR), where agents naturally fall back to editing state files when the expected CLI is unavailable. These results highlight a coverage boundary worth examining: auto mode assumes dangerous actions transit the shell, but agents routinely achieve equivalent effects through file edits that the classifier does not evaluate. --- ### Towards Predicting Multi-Vulnerability Attack Chains in Software Supply Chains from Software Bill of Materials Graphs Authors: Laura Baird, Armin Moin Published: 2026-04-04 Categories: cs.SE, cs.CR, cs.LG arXiv: https://arxiv.org/abs/2604.04977v2 Software supply chain security compromises often stem from cascaded interactions of vulnerabilities, for example, between multiple vulnerable components. Yet, Software Bill of Materials (SBOM)-based pipelines for security analysis typically treat scanner findings as independent per-CVE (Common Vulnerabilities and Exposures) records. We propose a new research direction based on learning multi-vulnerability attack chains through a novel SBOM-driven graph-learning approach. This treats SBOM structure and scanner outputs as a dependency-constrained evidence graph rather than a flat list of vulnerabilities. We represent vulnerability-enriched CycloneDX SBOMs as heterogeneous graphs whose nodes capture software components and known vulnerabilities (i.e, CVEs), connected by typed relations, such as dependency and vulnerability links. We train a Heterogeneous Graph Attention Network (HGAT) to predict whether a component is associated with at least one known vulnerability as a feasibility check for learning over this structure. Additionally, we frame the discovery of cascading vulnerabilities as CVE-pair link prediction using a lightweight Multi-Layer Perceptron (MLP) neural network trained on documented multi-vulnerability chains. Validated on 200 real-world SBOMs from the Wild SBOMs public dataset, the HGAT component classifier achieves 91.03% Accuracy and 74.02% F1-score, while the cascade predictor model (MLP) achieves a Receiver Operating Characteristic - Area Under Curve (ROC-AUC) of 0.93 on a seed set of 35 documented attack chains. --- ### StrADiff: A Structured Source-Wise Adaptive Diffusion Framework for Linear and Nonlinear Blind Source Separation Authors: Yuan-Hao Wei Published: 2026-04-04 Categories: stat.ML, cs.LG, cs.SD arXiv: https://arxiv.org/abs/2604.04973v2 This paper presents StrADiff, a Structured Source-Wise Adaptive Diffusion Framework for unsupervised blind source separation under linear and nonlinear mixing. The framework treats each latent dimension as a source branch and assigns to it an individual adaptive reverse diffusion mechanism, so that latent sources are recovered directly from observed mixtures through a single end-to-end objective, without supervised source labels or separate post-processing. Source-wise generation, structural regularization, and observation-space reconstruction are optimized jointly during training. In this instantiation, a Gaussian process (GP) prior is used as one example of a source-wise structured prior to impose temporal organization on each recovered trajectory; the framework itself is not restricted to GP priors and can in principle incorporate other structured priors. Theoretical components clarify the induced pushforward source law, the sample-level role of the structured prior, the coupling between source recovery and prior adaptation, and a conditional weak recovery statement in an idealized linear low-noise regime. Experiments on linear and nonlinear mixtures show that StrADiff can recover meaningful latent source trajectories in an unsupervised manner, with particularly stable performance in the linear case and moderate degradation under nonlinear mixing. Beyond classical signal separation, a source branch may also be interpreted as an independent, disentangled, or otherwise interpretable explanatory factor under suitable structural assumptions, suggesting a broader route toward structured latent modeling and future identifiable nonlinear representation learning. ---