# Research Papers — 2026-04 · bundle page 79 of 233 > Full-text markdown of papers #3901–3950 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/79 [← Previous bundle](https://aipod.dev/papers/archive/2026-04/page/78.md) [Next bundle →](https://aipod.dev/papers/archive/2026-04/page/80.md) ### Where Do Self-Supervised Speech Models Become Unfair? Authors: Felix Herron, Maja Hjuler, Solange Rossato, Alexandre Allauzen, François Portet Published: 2026-04-20 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.18249v1 Speech encoder models are known to model members of some speaker groups (SGs) better than others. However, there has been little work in establishing why this occurs on a technological level. To our knowledge, we present the first layerwise fairness analysis of pretrained self-supervised speech encoder models (S3Ms), probing each embedding layer for speaker identification (SID) automatic speech recognition (ASR). We find S3Ms produce embeddings biased against certain SGs for both tasks, starting at the very first latent layers. Furthermore, we find opposite patterns of layerwise bias for SID vs ASR for all models in our study: SID bias is minimized in layers that minimize overall SID error; on the other hand, ASR bias is maximized in layers that minimize overall ASR error. The inverse bias/error relationship for ASR is unaffected when probing S3Ms that are finetuned for ASR, suggesting SG-level bias is established during pretraining and is difficult to remove. --- ### Beyond Pattern Matching: Seven Cross-Domain Techniques for Prompt Injection Detection Authors: Thamilvendhan Munirathinam Published: 2026-04-20 Categories: cs.CR, cs.CL arXiv: https://arxiv.org/abs/2604.18248v4 Current open-source prompt-injection detectors converge on two architectural choices: regular-expression pattern matching and fine-tuned transformer classifiers. Both share failure modes recent work has made concrete. Regular expressions miss paraphrased attacks. Fine-tuned classifiers are vulnerable to adaptive adversaries: a 2025 NAACL Findings study reported that eight published indirect-injection defenses were bypassed with greater than fifty percent attack-success rates under adaptive attacks. This work proposes seven detection techniques that each port a mechanism from a discipline outside LLM security: forensic linguistics, materials-science fatigue analysis, deception technology, local-sequence alignment from bioinformatics, mechanism design, spectral signal analysis, and taint tracking. Each produces a signal architecturally independent of both regex matching and transformer classification, so the seven compose with existing defenses rather than replacing them. Four of seven are now implemented in prompt-shield v0.7.3 (Apache 2.0): d028 sequence alignment, d027 stylometric discontinuity, materials-fatigue tracking, and d034 honeypot tool definitions (new in v4.0). A four-configuration ablation across nine benchmarks (~10,300 samples) covers deepset, NotInject, LLMail-Inject, AgentHarm, AgentDojo, and an independent evaluation against three peer-reviewed academic benchmarks (Liu USENIX 2024, Garak, InjecAgent). This revision adds Section 5.7 (composed-stack adaptive-attack partial run, evidence for the composability thesis), Section 5.8 (50-document held-out benchmark: d027 collapses 1.000 to 0.000 F1 in isolation but the composed engine recovers 0.815 F1), and Section 7 (three architectural patterns extracted from prompt-shield in Gang-of-Four format). All code, data, and reproduction scripts are released Apache 2.0. --- ### Correction and Corruption: A Two-Rate View of Error Flow in LLM Protocols Authors: Fernando Reitich Published: 2026-04-20 Categories: cs.LG arXiv: https://arxiv.org/abs/2604.18245v3 Large language models operate in protocols containing multiple calls, yet added calls are usually evaluated only by their net effect. That summary cannot distinguish correcting unsuccessful outputs from corrupting initially successful ones. We develop a paired audit recording success before and after a specified operation on the same tasks under one binary rule. Correction and corruption rates exactly account for the net change: gains come from corrected failures and losses from corrupted successes. The break-even correction requirement rises sharply with baseline success, so the same behavior can improve a moderate-baseline population but harm a high-baseline one. Applied to published GPT-4 GSM8K results, the framework bounds within-sample correction and corruption counts from aggregate accuracies. We ask if calibration estimates predict unobserved outcomes, how rates change with information supplied to an operation, and whether successive measurements combine. Calibration estimates track accuracy on a disjoint sample from the same generator. Under reweighting of generator-defined groups, pooled estimates can fail while group-specific estimates reduce average prediction error. On GSM8K, observable features capture limited variation at these sample sizes, so the group-level application rule remains exploratory. Reordering a fixed four-candidate set changes both rates, with the accuracy effect depending on whether a correct alternative is available; on MBPP, adding a helper artifact raises corruptions among 357 initially passing programs from 28 to 100. Direct and composed transition estimates are close on average within samples; weighting initial-success rows by support substantially reduces held-out discrepancy. Correction and corruption are measurements of specified operations, not model constants, and can guide input choices, application decisions, and composition tests. --- ### Correction and Corruption: A Two-Rate View of Error Flow in LLM Protocols Authors: Fernando Reitich Published: 2026-04-20 Categories: cs.LG arXiv: https://arxiv.org/abs/2604.18245v2 Large language models are increasingly deployed as protocols: structured multi-call procedures that spend additional computation to transform a baseline answer into a final one. These protocols are evaluated only by end-to-end accuracy, giving limited insight into when they help, when they hurt, and whether their behavior transfers under distribution shift or composition. We propose a paired-outcome measurement interface for auditing a single protocol step on exact-match tasks. For each instance, the interface records a baseline correctness bit $E_0\in\{0,1\}$ and a post-step correctness bit $E_1\in\{0,1\}$, separating correction ($E_0=0\to E_1=1$) from corruption ($E_0=1\to E_1=0$) through two rates: $c=\Pr(E_1=1\mid E_0=0)$ and $γ=\Pr(E_1=0\mid E_0=1)$. These rates predict accuracy changes and define a reusable empirical interface testable across seeds, mixtures, and pipelines. We identify three failure mechanisms. Under mixture shift, pooled estimates of $(c,γ)$ become biased when calibration and deployment mixtures differ; conditioning on a difficulty proxy restores stability without additional model calls. Under presentation contamination, selection protocols alter the interface through stable presentation artifacts when candidate content is fixed. Under state insufficiency, the correctness bit may not carry enough history for multi-step pipelines to compose predictably; a Markov factorization test identifies when composition is valid and where additional state is needed. When a protocol step passes these diagnostics, it becomes an auditable module: gated by estimated gain, conditioned on a difficulty proxy to correct mixture bias, and composed into multi-step pipelines with predictable accuracy. We demonstrate these ideas on synthetic mathematical tasks and on GSM8K, where the calibrated interface correctly predicts when protocol steps should be activated or suppressed. --- ### Horospherical Depth and Busemann Median on Hadamard Manifolds Authors: Yangdi Jiang, Xiaotian Chang, Cyrus Mostajeran Published: 2026-04-20 Categories: math.ST, cs.LG, stat.ML arXiv: https://arxiv.org/abs/2604.18242v2 \We introduce the horospherical depth, an intrinsic notion of statistical depth on Hadamard manifolds, and define the Busemann median as the set of its maximizers. The construction exploits the fact that the linear functionals appearing in Tukey's half-space depth are themselves limits of renormalized distance functions; on a Hadamard manifold the same limiting procedure produces Busemann functions, whose sublevel sets are horoballs, the intrinsic replacements for halfspaces. The resulting depth is parametrized by the visual boundary, is isometry-equivariant, and requires neither tangent-space linearization nor a chosen base point. For arbitrary Hadamard manifolds, we prove that the depth regions are nested and geodesically convex, that a centerpoint of depth at least $1/(d+1)$ exists, and hence that the Busemann median exists for every Borel probability measure. Under strictly negative sectional curvature and mild regularity assumptions, the depth is strictly quasi-concave and the median is unique. We also establish robustness: the depth is stable under total-variation perturbations, and under contamination escaping to infinity the limiting median depends on the escape direction but not on how far the contaminating mass has moved along the geodesic ray, in contrast with the Fréchet mean. Finally, we establish uniform consistency of the sample depth and convergence of sample depth regions and sample Busemann medians; on symmetric spaces of noncompact type, the argument proceeds through a VC analysis of upper horospherical halfspaces, while on general Hadamard manifolds it follows from a compactness argument under a mild non-atomicity assumption. --- ### AJ-Bench: Benchmarking Agent-as-a-Judge for Environment-Aware Evaluation Authors: Wentao Shi, Yu Wang, Yuyang Zhao, Yuxin Chen, Fuli Feng, Xueyuan Hao, Xi Su, Qi Gu, Hui Su, Xunliang Cai, Xiangnan He Published: 2026-04-20 Categories: cs.AI arXiv: https://arxiv.org/abs/2604.18240v1 As reinforcement learning continues to scale the training of large language model-based agents, reliably verifying agent behaviors in complex environments has become increasingly challenging. Existing approaches rely on rule-based verifiers or LLM-as-a-Judge models, which struggle to generalize beyond narrow domains. Agent-as-a-Judge addresses this limitation by actively interacting with environments and tools to acquire verifiable evidence, yet its capabilities remain underexplored. We introduce a benchmark AJ-Bench to systematically evaluate Agent-as-a-Judge across three domains-search, data systems, and graphical user interfaces-comprising 155 tasks and 516 annotated trajectories. The benchmark comprehensively assesses judge agents' abilities in information acquisition, state verification, and process verification. Experiments demonstrate consistent performance gains over LLM-as-a-Judge baselines, while also revealing substantial open challenges in agent-based verification. Our data and code are available at https://aj-bench.github.io/. --- ### Towards Disentangled Preference Optimization Dynamics: Suppress the Loser, Preserve the Winner Authors: Wei Chen, Yubing Wu, Junmei Yang, Delu Zeng, Qibin Zhao, John Paisley, Min Chen, Zhou Wang Published: 2026-04-20 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2604.18239v4 Preference optimization is widely used to align large language models (LLMs) with human preferences. However, many margin-based methods also suppress the chosen response when they try to suppress the rejected one, and there is no general way to prevent this across different objectives. We address this issue with a unified incentive-score decomposition of preference optimization, revealing that different objectives share the same local update directions and differ only in their scalar weights. This decomposition provides a common framework for analyzing objectives that were previously studied in separate settings. Building on this decomposition, by analyzing the dynamics of the chosen/rejected likelihoods, we identify the disentanglement band (DB), a simple, testable condition that tells us when training can follow the desired path: suppress the loser while preserving the winner, possibly after an early stage. Using the DB, we propose reward calibration (RC), a plug-and-play method that adaptively rebalances the updates for chosen and rejected responses to satisfy the DB, without redesigning the base objective. Empirical results show that RC leads to more disentangled dynamics, with better downstream performance observed across several settings. Our code is available at https://github.com/IceyWuu/DisentangledPreferenceOptimization. --- ### Semantic-based Distributed Learning for Diverse and Discriminative Representations Authors: Zhuojun Tian, Chaouki Ben Issaid, Mehdi Bennis Published: 2026-04-20 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2604.18237v1 In large-scale distributed scenarios, increasingly complex tasks demand more intelligent collaboration across networks, requiring the joint extraction of structural representations from data samples. However, conventional task-specific approaches often result in nonstructural embeddings, leading to collapsed variability among data samples within the same class, particularly in classification tasks. To address this issue and fully leverage the intrinsic structure of data for downstream applications, we propose a novel distributed learning framework that ensures both diverse and discriminative representations. For independent and identically distributed (i.i.d.) data, we reformulate and decouple the global optimization function by introducing constraints on representation variance. The update rules are then derived and simplified using a primal-dual approach. For non-i.i.d. data distributions, we tackle the problem by clustering and virtually replicating nodes, allowing model updates within each cluster using block coordinate descent. In both cases, the resulting optimal solutions are theoretically proven to maintain discriminative and diverse properties, with a guaranteed convergence for i.i.d. conditions. Additionally, semantic information from representations is shared among nodes, reducing the need for common neural network architectures. Finally, extensive simulations on MNIST, CIFAR-10 and CIFAR-100 confirm the effectiveness of the proposed algorithms in capturing global structural representations. --- ### Negative Advantages Is a Double-Edged Sword: Calibrating advantages in GRPO for Search Agents Authors: Jiayi Wu, Ruobing Xie, Zeqian Huang, Lei Jiang, Can Xu, Kangyang Luo, Bochen Lin, Ming Gao, Xiang Li Published: 2026-04-20 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.18235v2 Search agents achieve strong question-answering performance through multi-turn interactions with search engines, with Group Relative Policy Optimization (GRPO) being a widely used training algorithm. However, GRPO-style algorithms still face several challenges in multi-hop search settings. First, correct intermediate steps are often penalized when the final answer is wrong. Second, training is highly unstable, often causing degradation of natural language ability or even catastrophic training collapse. Our analysis attributes these issues to coarse-grained advantage assignment and an imbalance between positive and negative advantages. To address these problems, we propose CalibAdv, an advantage calibration method specifically designed for search agents that enables more accurate and more stable modeling of penalties and rewards. Specifically, CalibAdv leverages the correctness of intermediate steps to downscale excessive negative advantages at a fine-grained level. It then further rebalances positive and negative advantages to improve training stability. Importantly, CalibAdv adopts a lightweight design that calibrates advantages from standard rollout signals, making it simple and easy to deploy. Extensive experiments across three models and seven benchmarks demonstrate that CalibAdv improves both model performance and training stability. Our code is available at https://github.com/wujwyi/CalibAdv. --- ### Evaluating Multi-Hop Reasoning in RAG Systems: A Comparison of LLM-Based Retriever Evaluation Strategies Authors: Lorenz Brehme, Thomas Ströhle, Ruth Breu Published: 2026-04-20 Categories: cs.IR, cs.AI arXiv: https://arxiv.org/abs/2604.18234v1 Retrieval-augmented generation (RAG) enhances large language models (LLMs) with external knowledge to answer questions more accurately. However, research on evaluating RAG systems-particularly the retriever component-remains limited, as most existing work focuses on single-context retrieval rather than multi-hop queries, where individual contexts may appear irrelevant in isolation but are essential when combined. In this research, we use the HotPotQA, MuSiQue, and SQuAD datasets to simulate a RAG system and compare three LLM-as-judge evaluation strategies, including our proposed Context-Aware Retriever Evaluation (CARE). Our goal is to better understand how multi-hop reasoning can be most effectively evaluated in RAG systems. Experiments with LLMs from OpenAI, Meta, and Google demonstrate that CARE consistently outperforms existing methods for evaluating multi-hop reasoning in RAG systems. The performance gains are most pronounced in models with larger parameter counts and longer context windows, while single-hop queries show minimal sensitivity to context-aware evaluation. Overall, the results highlight the critical role of context-aware evaluation in improving the reliability and accuracy of retrieval-augmented generation systems, particularly in complex query scenarios. To ensure reproducibility, we provide the complete data of our experiments at https://github.com/lorenzbrehme/CARE. --- ### Aether: Network Validation Using Agentic AI and Digital Twin Authors: Jordan Auge, Sam Betts, Giovanna Carofiglio, Giulio Grassi, Martin Gysi, John Kenneth d'Souza Published: 2026-04-20 Categories: cs.MA, cs.AI arXiv: https://arxiv.org/abs/2604.18233v1 Network change validation remains a critical yet predominantly manual, time-consuming, and error-prone process in modern network operations. While formal network verification has made substantial progress in proving correctness properties, it is typically applied in offline, pre-deployment settings and faces challenges in accommodating continuous changes and validating live production behavior. Current operational approaches typically involve scattered testing tools, resulting in partial coverage and errors that surface only after deployment. In this paper, we present Aether, a novel approach that integrates Generative Agentic AI with a multi-functional Network Digital Twin to automate and streamline network change validation workflows. It features an agentic architecture with five specialized Network Operations AI agents that collaboratively handle the change validation lifecycle from intent analysis to network verification and testing. Aether agents use a unified Network Digital Twin integrating modeling, simulation, and emulation to maintain a consistent, up-to-date network view for verification and testing. By orchestrating agent collaboration atop this digital twin, Aether enables automated, rapid network change validation while reducing manual effort, minimizing errors, and improving operational agility and cost-effectiveness. We evaluate Aether over synthetic network change scenarios covering main classes of network changes and on past incidents from a major ISP operational network, demonstrating promising results in error detection (100%), diagnostic coverage (92-96%), and speed (6-7 minutes) over traditional methods. --- ### FSEVAL: Feature Selection Evaluation Toolbox and Dashboard Authors: Muhammad Rajabinasab, Arthur Zimek Published: 2026-04-20 Categories: cs.LG arXiv: https://arxiv.org/abs/2604.18227v4 Feature selection is a fundamental machine learning and data mining task, involved with discriminating redundant features from informative ones. It is an attempt to address the curse of dimensionality by removing the redundant features, while unlike dimensionality reduction methods, preserving explainability. Feature selection is conducted in both supervised and unsupervised settings, with different evaluation metrics employed to determine which feature selection algorithm is the best. In this paper, we propose FSEVAL, a feature selection evaluation toolbox accompanied with a visualization dashboard, with the goal to make it easy to comprehensively evaluate feature selection algorithms. FSEVAL aims to provide a standardized, unified, evaluation and visualization toolbox to help the researchers working in the field, conduct extensive and comprehensive evaluation of feature selection algorithms with ease. --- ### FSEVAL: Feature Selection Evaluation Toolbox and Dashboard Authors: Muhammad Rajabinasab, Arthur Zimek Published: 2026-04-20 Categories: cs.LG arXiv: https://arxiv.org/abs/2604.18227v3 Feature selection is a fundamental machine learning and data mining task, involved with discriminating redundant features from informative ones. It is an attempt to address the curse of dimensionality by removing the redundant features, while unlike dimensionality reduction methods, preserving explainability. Feature selection is conducted in both supervised and unsupervised settings, with different evaluation metrics employed to determine which feature selection algorithm is the best. In this paper, we propose FSEVAL, a feature selection evaluation toolbox accompanied with a visualization dashboard, with the goal to make it easy to comprehensively evaluate feature selection algorithms. FSEVAL aims to provide a standardized, unified, evaluation and visualization toolbox to help the researchers working in the field, conduct extensive and comprehensive evaluation of feature selection algorithms with ease. --- ### Model in Distress: Sentiment Analysis on French Synthetic Social Media Authors: Pierre-Carl Langlais, Pavel Chizhov, Yannick Detrois, Carlos Rosas Hinostroza, Ivan P. Yamshchikov, Bastien Perroy Published: 2026-04-20 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.18226v1 Automated analysis of customer feedback on social media is hindered by three challenges: the high cost of annotated training data, the scarcity of evaluation sets, especially in multilingual settings, and privacy concerns that prevent data sharing and reproducibility. We address these issues by developing a generalizable synthetic data generation pipeline applied to a case study on customer distress detection in French public transportation. Our approach utilizes backtranslation with fine-tuned models to generate 1.7 million synthetic tweets from a small seed corpus, complemented by synthetic reasoning traces. We train 600M-parameter reasoners with English and French reasoning that achieve 77-79% accuracy on human-annotated evaluation data, matching or exceeding SOTA proprietary LLMs and specialized encoders. Beyond reducing annotation costs, our pipeline preserves privacy by eliminating the exposure of sensitive user data. Our methodology can be adopted for other use cases and languages. --- ### Is SAM3 ready for pathology segmentation? Authors: Qiuyu Kong, Shakiba Sharifi, Yiming Wang, Marco Cristani, Zanxi Ruan Published: 2026-04-20 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2604.18225v3 Is Segment Anything Model 3 (SAM3) capable in segmenting Any Pathology Images? Digital pathology segmentation spans tissue-level and nuclei-level scales, where traditional methods often suffer from high annotation costs and poor generalization. SAM3 introduces Promptable Concept Segmentation, offering a potential automated interface via text prompts. With this work, we propose a systematic evaluation protocol to explore the capability space of SAM3 in a structured manner. Specifically, we evaluate SAM3 under different supervision settings including zero-shot, few-shot, and supervised with varying prompting strategies. Our extensive evaluation on pathological datasets including NuInsSeg, PanNuke and GlaS, reveals that: (1) text-only prompts poorly activate nuclear concepts; (2) performance is highly sensitive to visual prompt types and budgets; (3) few-shot learning offers gains, but SAM3 lacks robustness against visual prompt noise; and (4) a significant gap persists between prompt-based usage and task-trained adapter-based reference. Our study delineates SAM3's boundaries in pathology image segmentation and provides practical guidance on the necessity of pathology domain adaptation. --- ### WebCompass: Towards Multimodal Web Coding Evaluation for Code Language Models Authors: Xinping Lei, Xinyu Che, Junqi Xiong, Chenchen Zhang, Yukai Huang, Chenyu Zhou, Haoyang Huang, Minghao Liu, Letian Zhu, Hongyi Ye, Jinhua Hao, Ken Deng et al. Published: 2026-04-20 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2604.18224v1 Large language models are rapidly evolving into interactive coding agents capable of end-to-end web coding, yet existing benchmarks evaluate only narrow slices of this capability, typically text-conditioned generation with static-correctness metrics, leaving visual fidelity, interaction quality, and codebase-level reasoning largely unmeasured. We introduce WebCompass, a multimodal benchmark that provides unified lifecycle evaluation of web engineering capability. Recognizing that real-world web coding is an iterative cycle of generation, editing, and repair, WebCompass spans three input modalities (text, image, video) and three task types (generation, editing, repair), yielding seven task categories that mirror professional workflows. Through a multi-stage, human-in-the-loop pipeline, we curate instances covering 15 generation domains, 16 editing operation types, and 11 repair defect types, each annotated at Easy/Medium/Hard levels. For evaluation, we adopt a checklist-guided LLM-as-a-Judge protocol for editing and repair, and propose a novel Agent-as-a-Judge paradigm for generation that autonomously executes generated websites in a real browser, explores interactive behaviors via the Model Context Protocol (MCP), and iteratively synthesizes targeted test cases, closely approximating human acceptance testing. We evaluate representative closed-source and open-source models and observe that: (1) closed-source models remain substantially stronger and more balanced; (2) editing and repair exhibit distinct difficulty profiles, with repair preserving interactivity better but remaining execution-challenging; (3) aesthetics is the most persistent bottleneck, especially for open-source models; and (4) framework choice materially affects outcomes, with Vue consistently challenging while React and Vanilla/HTML perform more strongly depending on task type. --- ### Instruction-as-State: Environment-Guided and State-Conditioned Semantic Understanding for Embodied Navigation Authors: Zhen Liu, Yuhan Liu, Jinjun Wang, Jianyi Liu, Wei Song, Jingwen Fu Published: 2026-04-20 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.18223v1 Vision-and-Language Navigation requires agents to follow natural-language instructions in visually changing environments. A central challenge is the dynamic entanglement between language and observations: the meaning of instruction shifts as the agent's field of view and spatial context evolve. However, many existing models encode the instruction as a static global representation, limiting their ability to adapt instruction meaning to the current visual context. We therefore model instruction understanding as an Instruction-as-State variable: a decision-relevant, token-level instruction state that evolves step by step conditioned on the agent's perceptual state, where the perceptual state denotes the observation-grounded navigation context at each step. To realize this principle, we introduce State-Entangled Environment-Guided Instruction Understanding (S-EGIU), a coarse-to-fine framework for state-conditioned segment activation and token-level semantic refinement. At the coarse level, S-EGIU activates the instruction segment whose semantics align with the current observation. At the fine level, it refines the activated segment through observation-guided token grounding and contextual modeling, sharpening its internal semantics under the current observation. Together, these stages maintain an instruction state that is continuously updated according to the agent's perceptual state during navigation. S-EGIU delivers strong performance on several key metrics, including a +2.68% SPL gain on REVERIE Test Unseen, and demonstrates consistent efficiency gains across multiple VLN benchmarks, underscoring the value of dynamic instruction--perception entanglement. --- ### EEG-Based Emergency Braking Intensity Prediction Using Blind Source Separation Authors: Zikun Zhou, Wenshuo Wang, Wenzhuo Liu, Hui Yao, Chaopeng Zhang, Yichen Liu, Xiaonan Yang, Junqiang Xi Published: 2026-04-20 Categories: cs.HC, cs.LG arXiv: https://arxiv.org/abs/2604.18220v1 Electroencephalography (EEG) signals have been promising for long-term braking intensity prediction but are prone to various artifacts that limit their reliability. Here, we propose a novel framework that models EEG signals as mixtures of independent blind sources and identifies those strongly correlated with braking action. Our method employs independent component analysis to decompose EEG into different components and combines time-frequency analysis with Pearson correlations to select braking-related components. Furthermore, we utilize hierarchical clustering to group braking-related components into two clusters, each characterized by a distinct spatial pattern. Additionally, these components exhibit trial-invariant temporal patterns and demonstrate stable and common neural signatures of the emergency braking process. Using power features from these components and historical braking data, we predict braking intensity at a 200 ms horizon. Evaluations on the open source dataset (O.D.) and human-in-the-loop simulation (H.S.) show that our method outperforms state-of-the-art approaches, achieving RMSE reductions of 8.0% (O.D.) and 23.8% (H.S.). --- ### Memorize When Needed: Decoupled Memory Control for Spatially Consistent Long-Horizon Video Generation Authors: Yanjun Guo, Zhengqiang Zhang, Pengfei Wang, Xinyue Liang, Zhiyuan Ma, Lei Zhang Published: 2026-04-20 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.18215v2 Spatially consistent long-horizon video generation aims to maintain temporal and spatial consistency along predefined camera trajectories. Existing methods mostly entangle memory modeling with video generation, leading to inconsistent content during scene revisits and diminished generative capacity when exploring novel regions, even trained on extensive annotated data. To address these limitations, we propose a decoupled framework that separates memory conditioning from generation. Our approach significantly reduces training costs while simultaneously enhancing spatial consistency and preserving the generative capacity for novel scene exploration. Specifically, we employ a lightweight, independent memory branch to learn precise spatial consistency from historical observation. We first introduce a hybrid memory representation to capture complementary temporal and spatial cues from generated frames, then leverage a per-frame cross-attention mechanism to ensure each frame is conditioned exclusively on the most spatially relevant historical information, which is injected into the generative model to ensure spatial consistency. When generating new scenes, a camera-aware gating mechanism is proposed to mediate the interaction between memory and generation modules, enabling memory conditioning only when meaningful historical references exist. Compared with the existing method, our method is highly data-efficient, yet the experiments demonstrate that our approach achieves state-of-the-art performance in terms of both visual quality and spatial consistency. --- ### TacticGen: Grounding Adaptable and Scalable Generation of Football Tactics Authors: Sheng Xu, Guiliang Liu, Tarak Kharrat, Yudong Luo, Mohamed Aloulou, Javier López Peña, Konstantin Sofeikov, Adam Reid, Paul Roberts, Steven Spencer, Joe Carnall, Ian McHale et al. Published: 2026-04-20 Categories: cs.AI, cs.LG, cs.MA arXiv: https://arxiv.org/abs/2604.18210v1 Success in association football relies on both individual skill and coordinated tactics. While recent advancements in spatio-temporal data and deep learning have enabled predictive analyses like trajectory forecasting, the development of tactical design remains limited. Bridging this gap is essential, as prediction reveals what is likely to occur, whereas tactic generation determines what should occur to achieve strategic objectives. In this work, we present TacticGen, a generative model for adaptable and scalable tactic generation. TacticGen formulates tactics as sequences of multi-agent movements and interactions conditioned on the game context. It employs a multi-agent diffusion transformer with agent-wise self-attention and context-aware cross-attention to capture cooperative and competitive dynamics among players and the ball. Trained with over 3.3 million events and 100 million tracking frames from top-tier leagues, TacticGen achieves state-of-the-art precision in predicting player trajectories. Building on it, TacticGen enables adaptable tactic generation tailored to diverse inference-time objectives through classifier guidance mechanism, specified via rules, natural language, or neural models. Its modeling performance is also inherently scalable. A case study with football experts confirms that TacticGen generates realistic, strategically valuable tactics, demonstrating its practical utility for tactical planning in professional football. The project page is available at: https://shengxu.net/TacticGen/. --- ### Towards Symmetry-sensitive Pose Estimation: A Rotation Representation for Symmetric Object Classes Authors: Andreas Kriegler, Csaba Beleznai, Margrit Gelautz Published: 2026-04-20 Categories: cs.CV, math.GT arXiv: https://arxiv.org/abs/2604.18208v1 Symmetric objects are common in daily life and industry, yet their inherent orientation ambiguities that impede the training of deep learning networks for pose estimation are rarely discussed in the literature. To cope with these ambiguities, existing solutions typically require the design of specific loss functions and network architectures or resort to symmetry-invariant evaluation metrics. In contrast, we focus on the numeric representation of the rotation itself, modifying trigonometric identities with the degrees of symmetry derived from the objects' shapes. We use our representation, SARR, to obtain canonic (symmetry-resolved) poses for the symmetric objects in two popular 6D pose estimation datasets, T-LESS and ITODD, where SARR is unique and continuous w.r.t. the visual appearance. This allows us to use a standard CNN for 3D orientation estimation whose performance is evaluated with the symmetry-sensitive cosine distance $\text{AR}_{\text{C}}$. Our networks outperform the state of the art using $\text{AR}_{\text{C}}$ and achieve satisfactory performance when using conventional symmetry-invariant measures. Our method does not require any 3D models but only depth, or, as part of an additional experiment, texture-less RGB/grayscale images as input. We also show that networks trained on SARR outperform the same networks trained on rotation matrices, Euler angles, quaternions, standard trigonometrics or the recently popular 6d representation -- even in inference scenarios where no prior knowledge of the objects' symmetry properties is available. Code and a visualization toolkit are available at https://github.com/akriegler/SARR . --- ### A Control Architecture for Training-Free Memory Use Authors: Yanzhen Lu, Muchen Jiang, Zhicheng Qian, Xingyu Zhou Published: 2026-04-20 Categories: cs.AI arXiv: https://arxiv.org/abs/2604.18206v1 Prompt-injected memory can improve reasoning without updating model weights, but it also creates a control problem: retrieved content helps only when it is applied in the right state. We study this problem in a strict training-free setting and formulate it as applicability control: when to trigger a memory-assisted second pass, when to trust it, and how to maintain the memory bank over time. Our method combines uncertainty-based routing, confidence-based selective acceptance, bank selection across rule and exemplar memory, and evidence-based governance of the memory bank over time. Under a locked training-free protocol with compute-matched controls, it improves two core arithmetic benchmarks by +7.0 points on SVAMP and +7.67 points on ASDiv over baseline. The same architecture also transfers to QA and agent benchmarks with smaller positive effects and shows the same positive direction on a second checkpoint for the main arithmetic tasks. On arithmetic, the main empirical pattern is that the control architecture, rather than raw memory exposure, drives the improvements on SVAMP and ASDiv. Mechanistically, confidence separates helpful from harmful rule-bank interventions, and under fixed retrieval the repair-versus-corrupt difference localizes to rows whose retrieved set actually contains the edited entries. --- ### A Comparative Evaluation of Geometric Accuracy in NeRF and Gaussian Splatting Authors: Mikolaj Zielinski, Eryk Vykysaly, Bartlomiej Biesiada, Jan Baturo, Mateusz Capala, Dominik Belter Published: 2026-04-20 Categories: cs.CV, cs.RO arXiv: https://arxiv.org/abs/2604.18205v1 Recent advances in neural rendering have introduced numerous 3D scene representations. Although standard computer vision metrics evaluate the visual quality of generated images, they often overlook the fidelity of surface geometry. This limitation is particularly critical in robotics, where accurate geometry is essential for tasks such as grasping and object manipulation. In this paper, we present an evaluation pipeline for neural rendering methods that focuses on geometric accuracy, along with a benchmark comprising 19 diverse scenes. Our approach enables a systematic assessment of reconstruction methods in terms of surface and shape fidelity, complementing traditional visual metrics. --- ### Hard to Be Heard: Phoneme-Level ASR Analysis of Phonologically Complex, Low-Resource Endangered Languages Authors: V. S. D. S. Mahesh Akavarapu, Michael Daniel, Gerhard Jäger Published: 2026-04-20 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.18204v1 We present a phoneme-level analysis of automatic speech recognition (ASR) for two low-resourced and phonologically complex East Caucasian languages, Archi and Rutul, based on curated and standardized speech-transcript resources totaling approximately 50 minutes and 1 hour 20 minutes of audio, respectively. Existing recordings and transcriptions are consolidated and processed into a form suitable for ASR training and evaluation. We evaluate several state-of-the-art audio and audio-language models, including wav2vec2, Whisper, and Qwen2-Audio. For wav2vec2, we introduce a language-specific phoneme vocabulary with heuristic output-layer initialization, which yields consistent improvements and achieves performance comparable to or exceeding Whisper in these extremely low-resource settings. Beyond standard word and character error rates, we conduct a detailed phoneme-level error analysis. We find that phoneme recognition accuracy strongly correlates with training frequency, exhibiting a characteristic sigmoid-shaped learning curve. For Archi, this relationship partially breaks for Whisper, pointing to model-specific generalization effects beyond what is predicted by training frequency. Overall, our results indicate that many errors attributed to phonological complexity are better explained by data scarcity. These findings demonstrate the value of phoneme-level evaluation for understanding ASR behavior in low-resource, typologically complex languages. --- ### Multiplication in Multimodal LLMs: Computation with Text, Image, and Audio Inputs Authors: Samuel G. Balter, Ethan Jerzak, Connor T. Jerzak Published: 2026-04-20 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.18203v1 Multimodal LLMs can accurately perceive numerical content across modalities yet fail to perform exact multi-digit multiplication when the identical underlying arithmetic problem is presented as numerals, number words, images, or in audio form. Because existing benchmarks often lack systematically paired instances across modalities, it remains difficult to compare genuine arithmetic limits within and across model families. We therefore introduce a controlled multimodal multiplication benchmark that factorially varies digit length, digit sparsity, representation (e.g., numerals vs. number words), and modality (text, rendered images, audio), with paired instances from a reproducible generator. We also define arithmetic load, C, as the product of the total and non-zero digit count as a compact, mechanistically motivated proxy for operation count. Across evaluations, accuracy falls sharply as C grows, often nearing zero by C > 100. Indeed, C remains predictive of performance across modalities and models, with R-squared often > 0.5, nearing the value from more complex measures of arithmetic load that count the number of intermediate arithmetic steps. A separate perception-versus-computation decomposition shows that multimodal degradation is primarily computational rather than perceptual: on matched-perception checks, models are near-perfect (> 99%) across modalities, even when multiplication accuracy drops. Beyond measuring when models fail, we ask which procedures they are predisposed to follow. We introduce a forced-completion loss probe that scores heuristic-specific reasoning prefixes--including columnar multiplication, distributive decomposition, and rounding/compensation. Here, decomposition is favored in both text and vision modalities; heuristic-specific LoRA adapters produce near-orthogonal updates yet degrade accuracy, indicating the base model maintains a well-tuned internal router. --- ### Centre manifold theorem for maps along manifolds of fixed points Authors: Lachlan Ewen MacDonald Published: 2026-04-20 Categories: math.DS, cs.LG arXiv: https://arxiv.org/abs/2604.18202v1 We prove a centre manifold theorem for a map along a manifold-with-boundary of fixed points, and provide an application to the study of gradient descent with large step size on two-layer matrix factorisation problems. --- ### DiffuSAM: Diffusion Guided Zero-Shot Object Grounding for Remote Sensing Imagery Authors: Geet Sethi, Panav Shah, Ashutosh Gandhe, Soumitra Darshan Nayak Published: 2026-04-20 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2604.18201v1 Diffusion models have emerged as powerful tools for a wide range of vision tasks, including text-guided image generation and editing. In this work, we explore their potential for object grounding in remote sensing imagery. We propose a hybrid pipeline that integrates diffusion-based localization cues with state-of-the-art segmentation models such as RemoteSAM and SAM3 to obtain more accurate bounding boxes. By leveraging the complementary strengths of generative diffusion models and foundational segmentation models, our approach enables robust and adaptive object localization across complex scenes. Experiments demonstrate that our pipeline significantly improves localization performance, achieving over a 14% increase in Acc@0.5 compared to existing state-of-the-art methods. --- ### Linear-Time and Constant-Memory Text Embeddings Based on Recurrent Language Models Authors: Tobias Grantner, Emanuel Sallinger, Martin Flechl Published: 2026-04-20 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.18199v1 Transformer-based embedding models suffer from quadratic computational and linear memory complexity, limiting their utility for long sequences. We propose recurrent architectures as an efficient alternative, introducing a vertically chunked inference strategy that enables fast embedding generation with memory usage that becomes constant in the input length once it exceeds the vertical chunk size. By fine-tuning Mamba2 models, we demonstrate their viability as general-purpose text embedders, achieving competitive performance across a range of benchmarks while maintaining a substantially smaller memory footprint compared to transformer-based counterparts. We empirically validate the applicability of our inference strategy to Mamba2, RWKV, and xLSTM models, confirming consistent runtime-memory trade-offs across architectures and establishing recurrent models as a compelling alternative to transformers for efficient embedding generation. --- ### Similarity-based Portfolio Construction for Black-box Optimization Authors: Catalin-Viorel Dinu, Diederick Vermetten, Carola Doerr Published: 2026-04-20 Categories: cs.NE arXiv: https://arxiv.org/abs/2604.18196v1 In black-box optimization, a central question is which algorithm to use to solve a given, previously unseen, problem. Selecting a single algorithm, however, entails inherent risks: inaccuracies in the selector may lead to poor choices, and even well-performing algorithms with high variance can yield unsatisfactory results in a single run. A natural remedy is to split the evaluation budget across multiple runs of potentially different algorithms. Such sequential algorithm portfolios benefit from variance reduction and complementarities between algorithms, often outperforming approaches that allocate the entire budget to a single solver. While effective portfolios can be constructed post-hoc, transferring this idea to the algorithm selection setting is non-trivial. We show that a naive portfolio constructed over the full training set already outperforms the strongest traditional baseline, the virtual best solver. We then propose a simple yet effective k-nearest-neighbor-based finetuning approach to construct portfolios tailored to unseen instances, yielding further improvements and highlighting the effectiveness of portfolio selection in fixed-budget black-box optimization. --- ### Friction-Augmented Drifting Models for Resource-Efficient Domain Translation Authors: Arkadii Kazanskii, Tatiana Petrova, Andrey Ustyuzhanin, Konstantin Bagrianskii, Aleksandr Puzikov, Radu State Published: 2026-04-20 Categories: cs.LG, cs.CV arXiv: https://arxiv.org/abs/2604.18194v2 Single-step generators promise high-fidelity synthesis at a fraction of the inference and training cost of ordinary differential equation (ODE)-based flow models, a central concern when compute is limited. Drifting Models (DMs) train a one-step generator by evolving samples under a kernel-based drift field, avoiding ODE integration entirely, but a two-particle surrogate of their iteration admits a \emph{locally repulsive} regime in which repulsion can dominate the attraction to the target. We introduce DMF (Drifting Model with Friction), which scales the drift field by a linearly-scheduled coefficient $1-γ(i)$. A closed-form analysis of the surrogate gives a per-step contraction threshold and a finite-horizon bound on the error trajectory, suggesting why friction can halt the iteration before it relaxes to a spurious force-balance fixed point. On FFHQ latent-space domain translation, DMF significantly improves on the frictionless DM it extends in both Fréchet Inception Distance (FID; paired $p=0.019$, Cohen's $d=1.71$) and CLIP-MMD (CMMD; $p=0.005$, $d=2.55$) with no additional forward passes or parameters, and on a 2D task it sharply improves DM's Fréchet (moment-matching) error while remaining on par under the 2-Wasserstein distance. DMF also achieves FID and CMMD comparable to the far more expensive Optimal Flow Matching (OFM) in our runs, at roughly $29\times$ lower training wall-clock on identical hardware. DMF thus delivers these gains with a single scheduled scalar. --- ### Attraction, Repulsion, and Friction: Introducing DMF, a Friction-Augmented Drifting Model Authors: Arkadii Kazanskii, Tatiana Petrova, Konstantin Bagrianskii, Aleksandr Puzikov, Radu State Published: 2026-04-20 Categories: cs.LG, cs.CV arXiv: https://arxiv.org/abs/2604.18194v1 Drifting Models (Deng et al., 2026) train a one-step generator by evolving samples under a kernel-based drift field, avoiding ODE integration at inference. The original analysis leaves two questions open. The drift-field iteration admits a locally repulsive regime in a two-particle surrogate, and vanishing of the drift ($V_{p,q}\equiv 0$) is not known to force the learned distribution $q$ to match the target $p$. We derive a contraction threshold for the surrogate and show that a linearly-scheduled friction coefficient gives a finite-horizon bound on the error trajectory. Under a Gaussian kernel we prove that the drift-field equilibrium is identifiable: vanishing of $V_{p,q}$ on any open set forces $q=p$, closing the converse of Proposition 3.1 of Deng et al. Our friction-augmented model, DMF (Drifting Model with Friction), matches or exceeds Optimal Flow Matching on FFHQ adult-to-child domain translation at 16x lower training compute. --- ### Scalable Neighborhood-Based Multi-Agent Actor-Critic Authors: Tim Goppelsroeder, Rasmus Jensen Published: 2026-04-20 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2604.18190v1 We propose MADDPG-K, a scalable extension to Multi-Agent Deep Deterministic Policy Gradient (MADDPG) that addresses the computational limitations of centralized critic approaches. Centralized critics, which condition on the observations and actions of all agents, have demonstrated significant performance gains in cooperative and competitive multi-agent settings. However, their critic networks grow linearly in input size with the number of agents, making them increasingly expensive to train at scale. MADDPG-K mitigates this by restricting each agent's critic to the $k$ closest agents under a chosen metric which in our case is Euclidean distance. This ensures a constant-size critic input regardless of the total agent count. We analyze the complexity of this approach, showing that the quadratic cost it retains arises from cheap scalar distance computations rather than the expensive neural network matrix multiplications that bottleneck standard MADDPG. We validate our method empirically across cooperative and adversarial environments from the Multi-Particle Environment suite, demonstrating competitive or superior performance compared to MADDPG, faster convergence in cooperative settings, and better runtime scaling as the number of agents grows. Our code is available at https://github.com/TimGop/MADDPG-K . --- ### Audio-DeepThinker: Progressive Reasoning-Aware Reinforcement Learning for High-Quality Chain-of-Thought Emergence in Audio Language Models Authors: Xiang He, Chenxing Li, Jinting Wang, Yan Rong, Tianxin Xie, Wenfu Wang, Li Liu, Dong Yu Published: 2026-04-20 Categories: cs.SD, cs.CL arXiv: https://arxiv.org/abs/2604.18187v1 Large Audio-Language Models (LALMs) have made significant progress in audio understanding, yet they primarily operate as perception-and-answer systems without explicit reasoning processes. Existing methods for enhancing audio reasoning rely either on supervised chain-of-thought (CoT) fine-tuning, which is limited by training data quality, or on reinforcement learning (RL) with coarse rewards that do not directly evaluate reasoning quality. As a result, the generated reasoning chains often appear well-structured yet lack specific acoustic grounding. We propose Audio-DeepThinker, a framework built on two core ideas. First, we introduce a hybrid reasoning similarity reward that directly supervises the quality of generated reasoning chains by combining an LLM evaluator assessing logical path alignment, key step coverage, and analytical depth with an embedding similarity component enforcing semantic alignment with reference reasoning chains. Second, we propose a progressive two-stage curriculum that enables high-quality CoT reasoning to emerge through pure RL exploration, without any supervised reasoning fine-tuning, from an instruction-tuned model that possesses no prior chain-of-thought capability. Stage 1 trains on foundational audio QA with the hybrid reward to foster basic reasoning patterns, while Stage 2 shifts to acoustically challenging boundary cases with an LLM-only reward for greater reasoning diversity. Audio-DeepThinker achieves state-of-the-art results on MMAR (74.0%), MMAU-test-mini (78.5%), and MMSU (77.26%), winning 1st Place in the Interspeech 2026 Audio Reasoning Challenge (Single Model Track). Interpretability analyses further reveal that RL training primarily reshapes upper-layer MoE gating mechanisms and that reasoning tokens crystallize progressively in the upper transformer layers, offering mechanistic insights into how audio reasoning emerges through exploration. --- ### Towards Multi-View Sign Language Understanding: A Benchmark Dataset and Baseline Authors: Xu Wang, Shengeng Tang, Wan Jiang, Yaxiong Wang, Lechao Cheng, Richang Hong Published: 2026-04-20 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.18184v2 Most existing Sign Language Understanding (SLU) methods are developed under fixed-view settings and remain vulnerable to viewpoint changes, while the scarcity of multi-view data hinders systematic research on viewpoint robustness. To address this gap, we introduce \textbf{MVSign}, a benchmark spanning three sign languages and seven viewpoints. MVSign combines original frontal recordings from existing datasets with additional views synthesized via $3$D whole-body motion reconstruction and target-view rendering, while preserving the source splits and annotations. We further propose \textbf{CanonSLU}, a framework for multi-view SLU. This framework employs a canonical-view-guided strategy, using the frontal view as a semantic anchor. Specifically, CanonSLU incorporates \emph{Canonical-View Knowledge Transfer (CKT)} to transfer canonical knowledge from the frontal view to non-frontal views, and \emph{Motion Relation Aggregation (MRA)} to aggregate motion-related features across frames and strengthen temporal representations under viewpoint changes. Extensive experiments on MVSign demonstrate competitive performance with gains on non-frontal views, establishing CanonSLU as an effective baseline for multi-view SLU. --- ### CanonSLR: Canonical-View Guided Multi-View Continuous Sign Language Recognition Authors: Xu Wang, Shengeng Tang, Wan Jiang, Yaxiong Wang, Lechao Cheng, Richang Hong Published: 2026-04-20 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.18184v1 Continuous Sign Language Recognition (CSLR) has achieved remarkable progress in recent years; however, most existing methods are developed under single-view settings and thus remain insufficiently robust to viewpoint variations in real-world scenarios. To address this limitation, we propose CanonSLR, a canonical-view guided framework for multi-view CSLR. Specifically, we introduce a frontal-view-anchored teacher-student learning strategy, in which a teacher network trained on frontal-view data provides canonical temporal supervision for a student network trained on all viewpoints. To further reduce cross-view semantic discrepancy, we propose Sequence-Level Soft-Target Distillation, which transfers structured temporal knowledge from the frontal view to non-frontal samples, thereby alleviating gloss boundary ambiguity and category confusion caused by occlusion and projection variation. In addition, we introduce Temporal Motion Relational Enhancement to explicitly model motion-aware temporal relations in high-level visual features, strengthening stable dynamic representations while suppressing viewpoint-sensitive appearance disturbances. To support multi-view CSLR research, we further develop a universal multi-view sign language data construction pipeline that transforms original single-view RGB videos into semantically consistent, temporally coherent, and viewpoint-controllable multi-view sign language videos. Based on this pipeline, we extend PHOENIX-2014T and CSL-Daily into two seven-view benchmarks, namely PT14-MV and CSL-MV, providing a new experimental foundation for multi-view CSLR. Extensive experiments on PT14-MV and CSL-MV demonstrate that CanonSLR consistently outperforms existing approaches under multi-view settings and exhibits stronger robustness, especially on challenging non-frontal views. --- ### Committed SAE-Feature Traces for Audited-Session Substitution Detection in Hosted LLMs Authors: Ziyang Liu Published: 2026-04-20 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2604.18179v3 Hosted-LLM providers have a silent-substitution incentive: advertise a stronger model while serving cheaper replies. Probe-after-return schemes such as SVIP leave a parallel-serve side-channel, since a dishonest provider can route the verifier's probe to the advertised model while serving ordinary users from a substitute. We propose a commit-open protocol that closes this gap. Before any opening request, the provider commits via a Merkle tree to a per-position sparse-autoencoder (SAE) feature-trace sketch of its served output at a published probe layer. A verifier opens random positions, scores them against a public named-circuit probe library calibrated with cross-backend noise, and decides with a fixed-threshold joint-consistency z-score rule. We instantiate the protocol on three backbones -- Qwen3-1.7B, Gemma-2-2B, and a 4.5x scale-up to Gemma-2-9B with a 131k-feature SAE. Of 17 attackers spanning same-family lifts, cross-family substitutes, and rank-<=128 adaptive LoRA, all are rejected at a shared, scale-stable threshold; the same attackers all evade a matched SVIP-style parallel-serve baseline. A white-box end-to-end attack that backpropagates through the frozen SAE encoder does not close the margin, and a feature-forgery attacker that never runs M_hon is bounded in closed form by an intrinsic-dimension argument. Commitment adds <=2.1% to forward-only wall-clock at batch 32. --- ### STaD: Scaffolded Task Design for Identifying Compositional Skill Gaps in LLMs Authors: Sungeun An, Swanand Ravindra Kadhe, Shailja Thakur, Chad DeLuca, Hima Patel Published: 2026-04-20 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.18177v2 Benchmarks are often used as a standard to understand LLM capabilities in different domains. However, aggregate benchmark scores provide limited insight into compositional skill gaps of LLMs and how to improve them. To make these weaknesses visible, we propose Scaffolded Task Design (STaD) framework. STaD generates controlled variations of benchmark tasks based on the concept of scaffolding, which introduces structured, incremental support in a step-by-step manner. Rather than inspecting failures individually, this approach enables systematic and scalable probing of model behavior by identifying the specific reasoning skill compositions they lack. Treating the LLM as a black box, our experiments on six models of varying sizes reveal multiple failure points in three reasoning benchmarks and highlight each model's unique and distinct skill gaps. --- ### QuantumQA: Enhancing Scientific Reasoning via Physics-Consistent Dataset and Verification-Aware Reinforcement Learning Authors: Songxin Qu, Tai-Ping Sun, Yun-Jie Wang, Huan-Yu Liu, Cheng Xue, Xiao-Fan Xu, Han Fang, Yang Yang, Yu-Chun Wu, Guo-Ping Guo, Zhao-Yun Chen Published: 2026-04-20 Categories: cs.AI, quant-ph arXiv: https://arxiv.org/abs/2604.18176v1 Large language models (LLMs) show strong capabilities in general reasoning but typically lack reliability in scientific domains like quantum mechanics, which demand strict adherence to physical constraints. This limitation arises from the scarcity of verifiable training resources and the inadequacy of coarse feedback signals in standard alignment paradigms. To address the data challenge, we introduce QuantumQA, a large-scale dataset constructed via a task-adaptive strategy and a hybrid verification protocol that combines deterministic solvers with semantic auditing to guarantee scientific rigor. Building on this foundation, we propose the verification-aware reward model (VRM) tailored for Reinforcement Learning with Verifiable Rewards (RLVR), which employs an adaptive reward fusion (ARF) mechanism to dynamically integrate deterministic signals from a scientific execution suite (SES) with multidimensional semantic evaluations for precise supervision. Experimental results demonstrate that our method consistently outperforms baselines and general-purpose preference models. Notably, our optimized 8B model achieves performance competitive with proprietary models, validating that incorporating verifiable, rule-based feedback into the reinforcement learning loop offers a parameter-efficient alternative to pure scaling. --- ### Copy-as-Decode: Grammar-Constrained Parallel Prefill for LLM Editing Authors: Ziyang Liu Published: 2026-04-20 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.18170v2 LLMs edit text and code by autoregressively regenerating the full output, even when most tokens appear verbatim in the input. We study Copy-as-Decode, a decoding-layer mechanism that recasts edit generation as structured decoding over a two-primitive grammar: <copy lines="i-j"/> references an input line range, <gen>...</gen> emits new content. A token-level FSM guarantees syntactic validity, and a serving-layer primitive updates the KV cache for each copy span via a single parallel-prefill forward rather than $N$ autoregressive steps -- sharing the parallel-forward kernel of speculative decoding but with input tokens as the draft and program-enforced acceptance replacing probabilistic verification. We report an upper-bound analysis that requires no end-to-end training. (i) Kernel speedup: on Qwen2.5-{1.5B, 7B}, copying $N$ tokens via parallel prefill is $6.8\times$--$303\times$ faster than autoregressive ($N \in (8, 512)$, A100 80GB bf16). (ii) Copy ceiling: on ProbeEdit and HumanEvalPack-Fix (Py/JS), $74$--$98\%$ of gold tokens are reachable under the line-level primitive; composed with the empirical kernel over each corpus's span histogram this yields a closed-form wall-clock bound of $29.0\times / 3.4\times / 4.2\times$ ($13.0\times$ pooled). A token-level extension reaches $91$--$99\%$ coverage with $4.5\times$--$6.5\times$ floors. (iii) Pipeline losslessness: oracle programs round-trip through the deterministic resolver on all $482$ cases, localizing any downstream failure to span selection rather than the mechanism. A perturbation study shows pooled EM drops from $100\%$ to $15.48\%$ under off-by-one noise. A fine-tuning pilot on Qwen2.5-Coder-1.5B lifts HEvalFix-Py EM from $0/33$ (untrained) to $12$--$17\%$, a learnability signal, not a production selector. Batched-serving integration and multi-file coverage are scoped as follow-up. --- ### Beyond Reproduction: A Paired-Task Framework for Assessing LLM Comprehension and Creativity in Literary Translation Authors: Ran Zhang, Steffen Eger, Arda Tezcan, Wei Zhao, Simone Paolo Ponzetto, Lieve Macken Published: 2026-04-20 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.18169v1 Large language models (LLMs) are increasingly used for creative tasks such as literary translation. Yet translational creativity remains underexplored and is rarely evaluated at scale, while source-text comprehension is typically studied in isolation, despite the fact that, in professional translation, comprehension and creativity are tightly intertwined. We address these gaps with a paired-task framework applied to literary excerpts from 11 books. Task 1 assesses source-text comprehension, and Task 2 evaluates translational creativity through Units of Creative Potential (UCPs), such as metaphors and wordplay. Using a scalable evaluation setup that combines expert human annotations with UCP-based automatic scoring, we benchmark 23 models and four creativity-oriented prompts. Our findings show that strong comprehension does not translate into human-level creativity: models often produce literal or contextually inappropriate renderings, with particularly large gaps for the more distant English-Chinese language pair. Creativity-oriented prompts yield only modest gains, and only one model, Mistral-Large, comes close to human-level creativity (0.167 vs. 0.246). Across all model-prompt combinations, only three exceed a creativity score of 0.1, while the rest remain at or near zero. --- ### Extending One-Step Image Generation from Class Labels to Text via Discriminative Text Representation Authors: Chenxi Zhao, Chen Zhu, Xiaokun Feng, Aiming Hao, Jiashu Zhu, Jiachen Lei, Jiahong Wu, Xiangxiang Chu, Jufeng Yang Published: 2026-04-20 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.18168v1 Few-step generation has been a long-standing goal, with recent one-step generation methods exemplified by MeanFlow achieving remarkable results. Existing research on MeanFlow primarily focuses on class-to-image generation. However, an intuitive yet unexplored direction is to extend the condition from fixed class labels to flexible text inputs, enabling richer content creation. Compared to the limited class labels, text conditions pose greater challenges to the model's understanding capability, necessitating the effective integration of powerful text encoders into the MeanFlow framework. Surprisingly, although incorporating text conditions appears straightforward, we find that integrating powerful LLM-based text encoders using conventional training strategies results in unsatisfactory performance. To uncover the underlying cause, we conduct detailed analyses and reveal that, due to the extremely limited number of refinement steps in the MeanFlow generation, such as only one step, the text feature representations are required to possess sufficiently high discriminability. This also explains why discrete and easily distinguishable class features perform well within the MeanFlow framework. Guided by these insights, we leverage a powerful LLM-based text encoder validated to possess the required semantic properties and adapt the MeanFlow generation process to this framework, resulting in efficient text-conditioned synthesis for the first time. Furthermore, we validate our approach on the widely used diffusion model, demonstrating significant generation performance improvements. We hope this work provides a general and practical reference for future research on text-conditioned MeanFlow generation. The code is available at https://github.com/AMAP-ML/EMF. --- ### Embedding Arithmetic: A Lightweight, Tuning-Free Framework for Post-hoc Bias Mitigation in Text-to-Image Models Authors: Venkatesh Thirugnana Sambandham, Torsten Schön Published: 2026-04-20 Categories: cs.CV arXiv: https://arxiv.org/abs/2604.18167v1 Modern text-to-image (T2I) models amplify harmful societal biases, challenging their ethical deployment. We introduce an inference-time method that reliably mitigates social bias while keeping prompt semantics and visual context (background, layout, and style) intact. This ensures context persistency and provides a controllable parameter to adjust mitigation strength, giving practitioners fine-grained control over fairness-coherence trade-offs. Using Embedding Arithmetic, we analyze how bias is structured in the embedding space and correct it without altering model weights, prompts, or datasets. Experiments on FLUX 1.0-Dev and Stable Diffusion 3.5-Large show that the conditional embedding space forms a complex, entangled manifold rather than a grid of disentangled concepts. To rigorously assess semantic preservation beyond the circularity and bias limitations of of CLIP scores, we propose the Concept Coherence Score (CCS). Evaluated against this robust metric, our lightweight, tuning-free method significantly outperforms existing baselines in improving diversity while maintaining high concept coherence, effectively resolving the critical fairness-coherence trade-off. By characterizing how models represent social concepts, we establish geometric understanding of latent space as a principled path toward more transparent, controllable, and fair image generation. --- ### MM-JudgeBias: A Benchmark for Evaluating Compositional Biases in MLLM-as-a-Judge Authors: Sua Lee, Sanghee Park, Jinbae Im Published: 2026-04-20 Categories: cs.CL, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2604.18164v3 Multimodal Large Language Models (MLLMs) have been increasingly used as automatic evaluators-a paradigm known as MLLM-as-a-Judge. However, their reliability and vulnerabilities to biases remain underexplored. We find that many MLLM judges fail to reliably integrate key visual or textual cues, yielding unreliable evaluations when evidence is missing or mismatched, and exhibiting instability under semantically irrelevant perturbations. To address this, we systematically define Compositional Bias in MLLM-as-a-Judge systems and introduce MM-JudgeBias, a benchmark for evaluating it. MM-JudgeBias introduces controlled perturbations across Query, Image, and Response, and evaluates model behavior via two complementary metrics: Bias-Deviation (BD) for sensitivity and Bias-Conformity (BC) for stability. Our dataset of over 1,800 curated and refined multimodal samples, drawn from 29 source benchmarks, enables a fine-grained diagnosis of nine bias types across diverse tasks and domains. Experiments on 26 state-of-the-art MLLMs reveal systematic modality neglect and asymmetric evaluation tendencies, underscoring the need for more reliable judges. --- ### Does "Do Differentiable Simulators Give Better Policy Gradients?'' Give Better Policy Gradients? Authors: Ku Onoda, Paavo Parmas, Manato Yaguchi, Yutaka Matsuo Published: 2026-04-20 Categories: cs.LG, cs.AI, cs.RO arXiv: https://arxiv.org/abs/2604.18161v1 In policy gradient reinforcement learning, access to a differentiable model enables 1st-order gradient estimation that accelerates learning compared to relying solely on derivative-free 0th-order estimators. However, discontinuous dynamics cause bias and undermine the effectiveness of 1st-order estimators. Prior work addressed this bias by constructing a confidence interval around the REINFORCE 0th-order gradient estimator and using these bounds to detect discontinuities. However, the REINFORCE estimator is notoriously noisy, and we find that this method requires task-specific hyperparameter tuning and has low sample efficiency. This paper asks whether such bias is the primary obstacle and what minimal fixes suffice. First, we re-examine standard discontinuous settings from prior work and introduce DDCG, a lightweight test that switches estimators in nonsmooth regions; with a single hyperparameter, DDCG achieves robust performance and remains reliable with small samples. Second, on differentiable robotics control tasks, we present IVW-H, a per-step inverse-variance implementation that stabilizes variance without explicit discontinuity detection and yields strong results. Together, these findings indicate that while estimator switching improves robustness in controlled studies, careful variance control often dominates in practical deployments. --- ### FreezeEmpath: Efficient Training for Empathetic Spoken Chatbots with Frozen LLMs Authors: Yun Hong, Yan Zhou, Yang Feng Published: 2026-04-20 Categories: cs.CL arXiv: https://arxiv.org/abs/2604.18159v1 Empathy is essential for fostering natural interactions in spoken dialogue systems, as it enables machines to recognize the emotional tone of human speech and deliver empathetic responses. Recent research has made significant progress in developing empathetic spoken chatbots based on large language models (LLMs). However, several challenges still exist when training such models, including reliance on costly empathetic speech instruction data and a lack of emotional expressiveness in the generated speech. Finetuning LLM with cross-modal empathetic instruction data may also lead to catastrophic forgetting and a degradation of its general capability. To address these challenges, we propose FreezeEmpath, an end-to-end empathetic spoken chatbot trained in a simple and efficient manner. The entire training process relies solely on existing speech instruction data and speech emotion recognition (SER) data, while keeping the LLM's parameters frozen. Experiments demonstrate that FreezeEmpath is able to generate emotionally expressive speech and outperforms other empathetic models in empathetic dialogue, SER, and SpokenQA tasks, demonstrating the effectiveness of our training strategy. --- ### State Transfer Reveals Reuse in Controlled Routing Authors: Yanzhen Lu, Zhicheng Qian, Muchen Jiang, Xingyu Zhou Published: 2026-04-20 Categories: cs.AI arXiv: https://arxiv.org/abs/2604.18158v1 Prompt-based interventions can change model behavior, but trained success alone does not identify where the behaviorally relevant state is represented. We study this question in controlled routing tasks using interfaces chosen on support data, held-out query evaluation, and matched necessity, sufficiency, and wrong-interface controls. On GPT-2 triop, an early interface supports exact transfer under these tests. On GPT-2 add/sub, zero-retrain compiled transfer at the fixed interface recovers most of donor routing accuracy, while trainable prompt slots can relearn the same behavior at several other positions only after additional support examples and optimization. These results distinguish fixed-interface reuse from prompt relocation in a setting where the two can be tested directly. Qwen routing provides a cross-architecture consistency check for the same matched-interface pattern at the operator token, although donor-specific identity on the local V-path remains unresolved. Generation and reasoning branches are used to map scope: they show broader transport or weaker controller identifiability once control depends on longer trajectories or harder selection. In controlled routing, fixed-interface transfer is therefore stronger evidence of reuse than trained prompt success alone. --- ### mlr3torch: A Deep Learning Framework in R based on mlr3 and torch Authors: Sebastian Fischer, Lukas Burk, Carson Zhang, Bernd Bischl, Martin Binder Published: 2026-04-20 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2604.18152v1 Deep learning (DL) has become a cornerstone of modern machine learning (ML) praxis. We introduce the R package mlr3torch, which is an extensible DL framework for the mlr3 ecosystem. It is built upon the torch package, and simplifies the definition, training, and evaluation of neural networks for both tabular data and generic tensors (e.g., images) for classification and regression. The package implements predefined architectures, and torch models can easily be converted to mlr3 learners. It also allows users to define neural networks as graphs. This representation is based on the graph language defined in mlr3pipelines and allows users to define the entire modeling workflow, including preprocessing, data augmentation, and network architecture, in a single graph. Through its integration into the mlr3 ecosystem, the package allows for convenient resampling, benchmarking, preprocessing, and more. We explain the package's design and features and show how to customize and extend it to new problems. Furthermore, we demonstrate the package's capabilities using three use cases, namely hyperparameter tuning, fine-tuning, and defining architectures for multimodal data. Finally, we present some runtime benchmarks. --- ### AI-based Waste Mapping for Addressing Climate-Exacerbated Flood Risk Authors: Steffen Knoblauch, Levi Szamek, Iddy Chazua, Benedcto Adamu, Innocent Maholi, Alexander Zipf Published: 2026-04-20 Categories: cs.CV, cs.CY arXiv: https://arxiv.org/abs/2604.18151v1 Urban flooding is a growing climate change-related hazard in rapidly expanding African cities, where inadequate waste management often blocks drainage systems and amplifies flood risks. This study introduces an AI-powered urban waste mapping workflow that leverages openly available aerial and street-view imagery to detect municipal solid waste at high resolution. Applied in Dar es Salaam, Tanzania, our approach reveals spatial waste patterns linked to informal settlements and socio-economic factors. Waste accumulation in waterways was found to be up to three times higher than in adjacent urban areas, highlighting critical hotspots for climate-exacerbated flooding. Unlike traditional manual mapping methods, this scalable AI approach allows city-wide monitoring and prioritization of interventions. Crucially, our collaboration with local partners ensured culturally and contextually relevant data labeling, reflecting real-world reuse practices for solid waste. The results offer actionable insights for urban planning, climate adaptation, and sustainable waste management in flood-prone urban areas. --- ### Attention-ResUNet for Automated Fetal Head Segmentation Authors: Ammar Bhilwarawala, Mainak Bandyopadhyay Published: 2026-04-20 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2604.18148v1 Automated fetal head segmentation in ultrasound images is critical for accurate biometric measurements in prenatal care. While existing deep learning approaches have achieved a reasonable performance, they struggle with issues like low contrast, noise, and complex anatomical boundaries which are inherent to ultrasound imaging. This paper presents Attention-ResUNet. It is a novel architecture that synergistically combines residual learning with multi-scale attention mechanisms in order to achieve enhanced fetal head segmentation. Our approach integrates attention gates at four decoder levels to focus selectively on anatomically relevant regions while suppressing the background noise, and complemented by residual connections which facilitates gradient flow and feature reuse. Extensive evaluation on the HC18 Challenge dataset where n = 200 demonstrates that Attention ResUNet achieves a superior performance with a mean Dice score of 99.30 +/- 0.14% against similar architectures. It significantly outperforms five baseline architectures including ResUNet (99.26%), Attention U-Net (98.79%), Swin U-Net (98.60%), Standard U-Net (98.58%), and U-Net++ (97.46%). Through statistical analysis we confirm highly significant improvements (p < 0.001) with effect sizes that range from 0.230 to 13.159 (Cohen's d). Using Saliency map analysis, we reveal that our architecture produces highly concentrated, anatomically consistent activation patterns, which demonstrate an enhanced interpretability which is crucial for clinical deployment. The proposed method establishes a new state of the art performance for automated fetal head segmentation whilst maintaining computational efficiency with 14.7M parameters and a 45 GFLOPs inference cost. Code repository: https://github.com/Ammar-ss --- ### The Magnitude of Dominated Sets: A Pareto Compliant Indicator Grounded in Metric Geometry Authors: Michael T. M. Emmerich Published: 2026-04-20 Categories: math.OC, cs.CG, cs.NE arXiv: https://arxiv.org/abs/2604.18147v1 We investigate \emph{magnitude} as a new unary and strictly Pareto-compliant quality indicator for finite approximation sets to the Pareto front in multiobjective optimization. Magnitude originates in enriched category theory and metric geometry, where it is a notion of size or point content for compact metric spaces and a generalization of cardinality. For dominated regions in the \(\ell_1\) box setting, magnitude is close to hypervolume but not identical: it contains the top-dimensional hypervolume term together with positive lower-dimensional projection and boundary contributions. This paper gives a first theoretical study of magnitude as an indicator. We consider multiobjective maximization with a common anchor point. For dominated sets generated by finite approximation sets, we derive an all-dimensional projection formula, prove weak and strict set monotonicity on finite unions of anchored boxes, and thereby obtain weak and strict Pareto compliance. Unlike hypervolume, magnitude assigns positive value to boundary points sharing one or more coordinates with the anchor point, even when their top-dimensional hypervolume contribution vanishes. We then formulate projected set-gradient methods and compare hypervolume and magnitude on biobjective and three-dimensional simplex examples. Numerically, magnitude favors boundary-including populations and, for suitable cardinalities, complete Das--Dennis grids, whereas hypervolume prefers more interior-filling configurations. Computationally, magnitude reduces to hypervolume on coordinate projections; for fixed dimension this yields the same asymptotic complexity up to a factor \(2^d-1\), and in dimensions two and three \(Θ(n\log n)\) time. These results identify magnitude as a mathematically natural and computationally viable alternative to hypervolume for finite Pareto front approximations. ---