# Research Papers — 2026-09 · bundle page 177 of 266 > Full-text markdown of papers #8801–8850 of 13298 published in 2026-09, newest first. AI-readable mirrors of the full 13298-paper 2026-09 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-09/page/177 [← Previous bundle](https://aipod.dev/papers/archive/2026-09/page/176.md) [Next bundle →](https://aipod.dev/papers/archive/2026-09/page/178.md) ### Partition-Aware Scheduling for Mobile Heterogeneous Inference Co-Execution Authors: Zhuojin Li, Marco Paolieri, Leana Golubchik Published: 2026-09-13 Categories: cs.DC, cs.LG, cs.PF arXiv: https://arxiv.org/abs/2609.14213v1 Modern mobile inference runs on heterogeneous platforms combining mobile GPUs with multiple CPU core clusters. Existing optimizations typically exploit either inter-operator parallelism, by assigning entire operators to CPU cores or to the GPU, or intra-operator parallelism, by partitioning each operator for CPU-GPU co-execution. We consider these two forms of parallelism together, to improve inference latency of tasks that can be represented by a static DAG of operators with predefined input/output tensor shapes (e.g., CNNs or vision transformers). We define the problem of partition-aware DAG scheduling for mobile heterogeneous inference, illustrating that the best strategy depends on the structure of the inference DAG, thus motivating a joint formulation capturing operator partition choices, device assignment, and execution order. We propose an online iterative search framework, which decomposes large DAGs into stages, focuses search on critical operators, and uses latency predictors to estimate partitioned execution without exhaustive profiling. Across representative mobile inference workloads, our approach achieves latency close to an offline solution while keeping scheduling overhead to a fraction of the model initialization cost, allowing platform-specific scheduling at deployment time. --- ### ECAS: An Edge-Controlled Agentic System for Validation-Gated Scientific Application Execution Authors: Baixi Sun, Mingze Xia, Huihuo Zheng Published: 2026-09-13 Categories: cs.DC, cs.AI arXiv: https://arxiv.org/abs/2609.14211v1 Scientific applications increasingly rely on high-performance computing (HPC), yet translating a scientist's high-level goal into a correct target-scale execution remains brittle and labor-intensive. Large language model (LLM) agents promise to automate this, but two obstacles remain: granting a cloud-hosted model direct HPC access exposes credentials and execution authority, while withholding it demands continuous human supervision; and one-shot generation cannot adapt when generated artifacts fail in a site-specific HPC environment. We present \textsc{ECAS}, an \textbf{E}dge-\textbf{C}ontrolled \textbf{A}gentic \textbf{S}ystem for closed-loop execution of scientific computing campaigns with limited human intervention. \textsc{ECAS} separates \emph{reasoning}, \emph{control}, and \emph{execution}: a cloud-hosted LLM proposes plans, artifacts, and repairs; a user-controlled edge agent retains credentials, workflow state, and execution authority while enforcing policy and resource constraints; and the HPC system computes. Its core mechanism is \emph{validation-gated execution}: generated artifacts pass static checks and small-scale validation, failures trigger repairs from sanitized execution feedback, and target-scale execution is permitted only after validation and policy checks pass. \textsc{ECAS} also draws on an edge-resident library of expert-distilled, site-specific skills that is never disclosed to the cloud. In preliminary experiments with three scientific applications on two production ALCF systems under six injected fault types, closed-loop repair improves application success from 0/6 to 6/6 over one-shot generation, validation gating prevents all three observed target-scale failures, and skill conditioning improves success from 4/6 to 6/6. These results show the feasibility of delegating adaptive reasoning to the cloud while retaining execution control at the edge. --- ### Learning to Refer from Estimated Listener Gaze Authors: Téa Wright, Alane Suhr Published: 2026-09-13 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.14207v1 We propose to finetune vision-language models to generate more pragmatically optimal referring expressions by transforming observations of incremental listener comprehension, in the form of gaze scanpaths, into learning signals. During training, referring expressions are sampled from the speaker policy being optimized, conditioned on images and target referents; then, a neural listener estimating human gaze behavior maps from images and sampled referring expressions to scanpaths, each represented by a sequence of fixations, with each fixation corresponding to a word in the referring expression. We experiment with several approaches to convert fixation sequences and target referents into token- and sequence-level rewards, which are used to optimize policy parameters. Through evaluation with human listeners, we find that speaker policies trained with gaze-estimating listeners result in significantly more pragmatically-optimal references than base models, reducing sequence length from 15.4 down to 4.0 words while increasing referential success from 75.2 up to 80.0%. Our work demonstrates a promising opportunity for learning to generate utterances through language-based interaction, not only from the explicit signal of communicative success, but also from implicitly-available observations of a listener's process of comprehension. --- ### Typed Temporal Interaction Features for Simulation-Backed Forecasting of Open-Source Game Release Incidents Authors: Shayma Alkobaisi, Anas Ali Published: 2026-09-12 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.31647v1 Open-source video-game quality depends on inter-actions among code, assets, configuration, tests, contributors, and issue workflows, yet conventional defect predictors usually flatten or omit these relations. We investigate release-level forecasting of a quality incident within thirty days using GAMEQUALGRAPH-Pilot, a typed temporal feature pipeline with calibrated risk estimates and effort-aware ranking. Because the accessible OS-SGameBench materials do not provide manually audited release dates and outbreak labels, the executed evaluation is explicitly simulation-backed rather than an empirical claim about real games. Five seeded worlds each contain 120 projects and 24 releases, with project-disjoint validation and future cross-project testing. The pilot obtains an AUPRC of 0.520, AUROC of 0.673, Brier score of 0.207, and 29.68% effort-aware recall at a twenty-percent testing budget. Its closest local comparator, Static-Hetero-Reimpl, reaches 0.522 AUPRC; the -0.002 difference is not statistically significant after Holm correction. Inference requires 0.023 milliseconds per release in the measured environment. Ablations and controlled missingness, drift, engine, project-size, alert-threshold, and attribution analyses expose where typed interactions help and where they fail. Results support the reproducibility of the proposed protocol, not deployment effectiveness. Real OSSGameBench release reconstruction, stratified label audits, and official graph-model comparisons remain mandatory before journal submission or operational use in practice. This boundary protects research integrity and supports credible evaluation. --- ### Measurement-Error-Aware Causal Distributed-Lag Quantile Modeling of Indoor Air Pollution and Short-Term Lung-Function Deterioration Authors: Shayma Alkobaisi, Anas Ali Published: 2026-09-12 Categories: stat.AP, cs.AI, cs.CY, cs.LG arXiv: https://arxiv.org/abs/2609.31646v1 Low-cost indoor air-quality sensors could support personalized asthma prevention, but their nonlinear measurement error, delayed exposure effects, time-varying confounding, and heterogeneous lower-tail responses limit risk estimation. We present CAUSALQUANT-ASTHMA, a measurement-error-aware causal quantile distributed-lag framework for short-horizon peak expiratory flow analysis. Sparse reference measurements train a nonlinear calibration model; stabilized sequential generalized-propensity weights address measured exposure assignment; and a susceptibility-modulated, smooth, noncrossing quantile model estimates lag-specific and sustained-exposure contrasts. Because no authorized cohort simultaneously provided dense indoor sensing, reference co-location, and outcome-compatible longitudinal data, evaluation used five semi-synthetic panels with known counterfactual truth, 150 patients and 12,600 patient-days per realization. Across eight methods, CAUSALQUANT achieved a dose-response integrated absolute error of 0.304 plus or minus 0.094, improving 24.2 percent over the strongest measurement-error and propensity-weighted baseline. It also obtained the lowest overall pinball loss, 1.065, while maintaining zero quantile crossings and 78.1 percent coverage for the nominal 80 percent interval. Sensor calibration reduced held-out exposure RMSE by 33.7 percent. Stress tests quantified degradation under sensor noise, missing personal measurements, and hidden confounding. These findings establish methodological feasibility and reproducibility, not clinical effectiveness; prospective, governance-approved external validation is required before patient-level interpretation or deployment. --- ### STAR: Adaptive Spatial-Temporal Normalization for Unified Microservice Incident Management Authors: Xinhua Miao, Linyu Zhu, Bowei Yang, Zhengong Cai Published: 2026-09-12 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.31645v1 Automated incident management in large-scale microservice systems relies on learning robust representations from multimodal observability data, including metrics, logs, and traces. Although recent self-supervised frameworks enable unified modeling for anomaly detection (AD), failure triage (FT), and root cause localization (RCL), they often struggle with non-stationary temporal dynamics and heterogeneous service dependency structures. In this paper, we propose STAR, a Spatial-Temporal Adaptive Representation learning framework that explicitly addresses these challenges through adaptive normalizations. STAR introduces two tightly coupled mechanisms: Temporal Adaptive Normalization (TAN), which dynamically normalizes multivariate time series using multi-scale temporal context, and Spatial Adaptive Normalization (SAN), which performs structure-aware normalization over service dependency graphs. Unlike prior methods that treat normalization as static or task-agnostic, STAR formulates it as a learnable, context-conditioned transformation aligned with the intrinsic properties of microservice systems. The resulting adaptive representations are integrated into a unified self-supervised framework, enabling end-to-end unsupervised support for AD, FT, and RCL tasks. Extensive experiments on two real-world microservice benchmarks demonstrate that STAR consistently outperforms all state-of-the-art baselines, yielding significant and stable improvements across all three tasks. Our results highlight adaptive normalization as a principled and effective mechanism for robust multimodal representation learning in complex software systems. --- ### MaD-RL: Matching Distributions for Calibrating LLMs with Reinforcement Learning Authors: Sourabh Kulkarni, Ksheeraj Sai Vepuri, Basar Demir, Jason Bohrer, Emily Shen, Jianfa Chen, Nan Jiang, Ankit Jain, Harihar Subramanyam, Mannat Singh, Chirag Nagpal Published: 2026-09-12 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.31644v1 Reinforcement learning (RL) is widely used in language-model post-training to maximize rewards assigned to individual model outputs, such as scores from binary verifiers or reward models trained on human feedback. However, applications such as synthetic-data generation, fairness-related constraint satisfaction, and policy exploration require controlling the distribution of outputs across model generations rather than only maximizing expected reward. We propose a general RL-based framework for \textit{Distribution Matching} allowing matching the distribution of a latent categorical attribute of model outputs to a specified target distribution. Empirically, we demonstrate that dominant post-training recipes such as Group Relative Policy Optimization (GRPO) reduce output diversity by concentrating policy probability towards a single mode. Entropy regularization and sampling temperature can improve the spread of the distribution but have constrained effectiveness, limited to apply only in token space and toward uniform distributions. We show that prior work in this area is a specific case of Distribution Matching involving the $L_2$ divergence. We then propose reward functions for other divergences such as KL and Jensen-Shannon and motivate them with theoretical justification. Finally, we demonstrate the effectiveness of our approach on a set of experiments involving mathematical reasoning and programming. --- ### Information Design Against Gaming and Learning Adversaries Authors: Madhava Gaikwad Published: 2026-09-12 Categories: cs.LG, cs.AI, cs.CR, cs.GT arXiv: https://arxiv.org/abs/2609.31643v1 A principal who deploys a binary classifier with an abstention option must decide which queries the mechanism abstains on. The right choice depends on the adversary. A gaming adversary already knows the classifier and tries to manipulate features across the boundary, so the principal does best by abstaining on queries close to that boundary. The same boundary-localizing rule is the worst possible choice against a learning adversary who does not know the classifier: each abstention now tells the adversary that the boundary is nearby, which is enough to drive a binary search. We analyze this tension. The two natural defenses, abstaining at a fixed rate and abstaining near the boundary, are Blackwell-incomparable: neither can be simulated by post-processing the other's responses. The number of queries needed to reconstruct the boundary to error $\eps$ is $\tildeΘ(d/\eps)$ under the first defense and $Θ(d \log(1/\eps))$ under the second, where $d$ is the VC dimension of the classifier family and $\tildeΘ$ suppresses factors polylogarithmic in $d$ and $1/\eps$. The first rate is a worst case over query distributions; no reconstruction algorithm can close the gap at the distributions that attain it. We characterize the Pareto frontier between the two defense objectives, and confirm both rates on seven binary-classification tasks spanning tabular, image, and language-model-feature inputs: label-plus-counterfactual access extracts the boundary with up to $200\times$ fewer queries than a published label-only baseline. --- ### A Survey on Fake Review Detection: From Pre-trained Language Models to Large Language Models Authors: Fanji Yang, Huiyao Chen, Xi Yu, Meishan Zhang, Xiaohong Xiao, Mingsen Deng Published: 2026-09-12 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2609.30292v1 Online reviews shape consumer decisions, platform governance, and corporate reputation.Fake reviews compromise this information channel by injecting deceptive evidence into rating systems, recommendation pipelines, and public trust mechanisms.The rise of large language models, or LLMs, has changed the problem in two directions.LLMs can generate fluent and context-aware deceptive reviews, while pre-trained language models, or PLMs, and LLMs also provide stronger semantic representations for detection.This survey reviews fake review detection from an information fusion perspective, covering 211 studies published from 2018 to early 2026.We organize existing work by evidence source and fusion level, covering review text, sentiment, rating behavior, temporal metadata, user-product graphs, multimodal content, external knowledge, and LLM-generated signals.We trace the development from traditional machine learning and deep learning to PLM-based and LLM-based methods, and examine how different approaches combine textual, behavioral, structural, and multimodal evidence.We also analyze reported performance trends on widely used Amazon, Yelp, and OpSpam benchmark families, while noting the limitations caused by different label construction procedures, data splits, and evaluation protocols.Finally, we identify open problems in adversarial generation, cross-domain transfer, uncertainty-aware fusion, missing-source robustness, interpretability, and trustworthy evaluation for AI-generated deceptive content. --- ### LLMersion: A Local-First AI Agent Framework for Low-Cost Home Language Learning toward Educational Equity Authors: Qiming Guo, Jinwen Tang, Xingran Huang, Hung-Yu Lin, Yafu Zhong, Xiatian Zhuang Published: 2026-09-12 Categories: cs.CL, cs.CY, cs.HC arXiv: https://arxiv.org/abs/2609.29672v1 Artificial intelligence helps education most where an essential provision has been rationed by cost. For language learners that provision is a teacher's voice, which binds listening, reading, speaking, and writing into one act. Published evidence shows why most learners lack it, from a global shortage of 44 million teachers to heavy household tutoring bills, and why technology has not substituted for it: computer-assisted language learning proved effective but narrow, applications presuppose connectivity 2.6 billion people lack, and One Laptop per Child's randomized evaluation found that hardware without capable software teaches nothing. We distill eight difficulties and four binding constraints, and argue that small open-weight models dissolve the last: a complete four-skill stack now fits a \$200-class laptop and, on community measurements, generates at the pace speech is consumed, for about one US cent of electricity per study hour. We therefore propose LLMersion, a scheme for AI for education that runs entirely at home, over the learner's own documents, with an AI-written, AI-understood, AI-updated codebase anyone can customize; present LLMersion-1, a released open-source prototype (https://github.com/QM378/LLMersion); and outline the vision of a private learning agent. --- ### An Affordable AI-Integrated Smart Cane for Multimodal Mobility Assistance of Visually Impaired Users Authors: Ali Akarma, Adeel Ahmad, Toqeer Ali Syed Published: 2026-09-12 Categories: cs.AI, cs.CV, cs.CY, cs.RO arXiv: https://arxiv.org/abs/2609.22277v1 Visual impairment affects over 2.2 billion people worldwide, yet conventional white canes cannot detect elevated hazards or provide semantic environmental context. Existing AI-assisted navigation systems typically rely on expensive hardware or cloud connectivity, limiting accessibility in resource-constrained settings. This paper presents an affordable (\$88 USD), fully offline AI-integrated smart cane designed for multimodal mobility assistance on an ultra-low-power Raspberry Pi Zero 2W. The system fuses RGB vision sensing with Time-of-Flight (ToF) distance estimation, pairing an INT8-quantized SSD MobileNet V1 model with distance-aware vibrotactile feedback and real-time audio alerts. To ensure operational robustness on constrained hardware, a multiprocessing architecture isolates sensor acquisition, neural inference, and haptic feedback into independent processes with fail-safe sensing support. Experimental evaluation across indoor mobility scenarios demonstrates a macro-averaged F1-score of 0.82 (precision: 0.85, recall: 0.81), a mean end-to-end latency of 330\,ms, and a peak power draw of 2.8\,W. A preliminary usability study with 12 participants (SUS: 78.5, NASA-TLX) demonstrated positive user perception and enhanced obstacle awareness. The proposed prototype validates the feasibility of deploying privacy-preserving, edge-native assistive intelligence for cost-sensitive mobility assistance. --- ### Prior-Free Competitive Ratios for Improving Bandits: Scale, Curvature and Horizon Are Free, but Not Jointly Under Noise Authors: Xuan Li Published: 2026-09-12 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.17595v1 In the improving multi-armed bandits problem, each of $k$ arms has an unknown nondecreasing, discretely concave reward curve $f_i$, and pulling arm $i$ for the $t$-th time yields $f_i(t)$. For sufficiently long horizons, Blum and Ravichandran (ALT 2025) proved that randomized algorithms achieve an $O(\sqrt k)$ approximation to the best single arm when the scale $m=f^*(T)$ of the optimal arm is known ($T\ge2k$), and $O(\sqrt k\log k)$ when it is not ($T>4k$), against an $Ω(\sqrt k)$ lower bound. The logarithmic factor is unnecessary: a one-page \emph{probe-and-commit} algorithm achieves competitive ratio $4\sqrt3\,\sqrt k$ for $T\ge2\lfloor\sqrt k\rfloor$, without any knowledge of the scale, and we determine the optimal ratio for every horizon, $Θ(\sqrt k+k/T)$, also for unknown horizons. Without noise, \emph{no prior is needed at all}: a random-marginal probing algorithm reading neither the scale $m$, nor the concavity-envelope exponent $β$ of Blum, Garicano, Ravichandran and Sharma (UAI 2026), nor the horizon $T$, achieves the optimal $Θ(k^{β/(1+β)}+k/T)$ simultaneously for every $β$ and every horizon. Under the multiplicative noise model of Blum and Ravichandran, probe-and-commit keeps the same all-horizon order $Θ(\sqrt k+k/T)$ without knowing the noise level (and $Θ(\sqrt k)$ on the same range), but the price of priors jumps: for any fixed noise level $\varepsilon\in(0,1/2)$, the uniform price of adaptation $φ_\varepsilon(k)$ --- the worst case over horizons $T\ge16k$ of the loss relative to $k^{β/(1+β)}$ for algorithms knowing neither $m$ nor $β$ --- is $Θ_\varepsilon(\sqrt{\log k/\log\log k})$, the lower bound asymptotic in $k$ at fixed positive $\varepsilon$ and matched by a nested random-permutation probing algorithm, whereas knowing either $m$ or $β$ alone restores a constant price. --- ### When the Gradient Sees Rank: Provable Necessity, Causal Recruitment, and Composition in Trained Matrix Memories Authors: Samuel Larson Published: 2026-09-12 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.17594v1 Can gradient-based training learn the rank needed to store and compose associations in a matrix memory? In our earlier study, we used a matrix-augmented reasoner on a task that admits a rank-1 solution, leaving this question open. We train matrix memories on $K$ fresh key-value bindings whose exact linear recovery requires $\mathrm{rank}(Z) \geq K$. A fixed linear readout queries a single matrix state without access to the original bindings. Experiments measure recovery by cosine similarity greater than 0.9, a threshold distinct from mathematical equality. Learned effective rank increases with $K$ across the tested grid (Spearman $ρ= 1.0$ at $d = 16$). Training-time rank caps produce a recovery transition near $k = K$: at $d = 8$, $K = 4$, rank 3 gives at most 0.0004 recovery and rank 4 gives 0.97. Four of five seeds retain at least 0.9996 recovery through 21-fold self-application of the trained operator. On the entity subspace, the learned operator has effective rank close to $K$ and approximates the ideal cycle. For the single converged seed capped below $K$, a calculation using the entity-subspace operator and ideal cycle predicts the measured cosine within 0.008 through seven applications. Extending training resolves several initial failures, but recovery still declines at larger matrix dimensions with encoder width fixed. --- ### Data-free On-policy Distillation Authors: Gengsheng Li, Mao Zheng, Mingyang Song, Jie Sun, Zeyuan Liu, Ruiqi Liu, Qiyong Zhong, Haiyun Guo, Junfeng Fang, Jinqiao Wang Published: 2026-09-12 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.14193v2 On-policy distillation (OPD) has become a standard component of frontier post-training pipelines, yet how much its training data actually contributes has gone largely unexamined. On the two teacher--student pairings most common in practice, we find OPD almost indifferent to its data: eight prompts already match a 17k-problem dataset, and three independently built datasets whose difficulty and teacher--student KL differ several-fold produce nearly indistinguishable training curves. Two causes account for this. First, the unit of data in OPD is the state a prompt leads to, not the prompt itself: a single prompt keeps exposing new teacher correction as sampling continues, while the marginal value of additional prompts collapses after eight. Second, replacing mathematics with competitive programming still recovers over ninety percent of the in-domain gain, indicating that OPD transfers the teacher's mode of reasoning rather than knowledge related to the data. We take this to its limit with \textbf{Data-free On-policy Distillation} (DF-OPD), in which the teacher writes its own training questions under a simple prompt---no external data, no quality filtering---leaving a system of just two policies. DF-OPD matches and even surpasses real data, and the questions it produces track the teacher's own post-training data on three key diagnostics of training dynamics, which other real datasets do not. Applied to multi-teacher distillation, where the (prompt, domain) pairs normally have to be derived from post-training data that is often out of reach, 1k self-generated questions close 98.5\% of the available headroom, even surpassing the 96.6\% reached with 7k real examples. Together these results invite a reassessment of the role data plays in OPD. --- ### MorphoStyle: Motion Style Transfer with Morphology Control Authors: Xin Feng, Eleonora D'Arnese, Mohan Sridharan Published: 2026-09-12 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.14189v1 Human motion may be viewed as a combination of action content, style, and body morphology. Existing motion style transfer methods transfer a reference style onto a content motion while assuming a canonical body, whereas shape-aware motion generators adapt motion to a target shape without explicit style control. This separation of motion style and shape (morphology) makes it difficult to generate stylized motions for non-canonical bodies; naively combining a style transfer module with a shape-aware generator often leaks action content from the style reference and disrupts shape-consistent kinematics. We present MorphoStyle, a framework for shape-aware motion style transfer that is built on a shape-conditioned Finite-Scalar-Quantization Variational Auto-Encoder (FSQ-VAE). The key contribution is to pose the desired style transfer as modular latent disentanglement comprising: (i) a contrastive style encoder that extracts content-decoupled style embeddings; (ii) a text-guided style-routing mechanism that locates style-relevant joints in a text-motion feature space; and (iii) a manifold preserving style modulator that injects discriminative style embeddings in content features as a temporally-gated low-rank offset. Extensive experiments on benchmark datasets demonstrate that MorphoStyle outperforms competing baselines in terms of both shape control and motion style transfer, while simultaneously providing quantitative shape control. For more details, please see project website: https://github.com/funkdub/MorphoStyle. --- ### Entropy-Punctured Bloom Filters for Memory-Efficient Machine Learning Authors: John Cartmell, Mihaela Cardei, Ionut Cardei Published: 2026-09-12 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.14187v1 Memory-efficient feature representations are increasingly important in machine learning settings where storage, transmission cost, bandwidth, or privacy constraints limit access to raw data. Bloom Filter (BF) encodings provide compact probabilistic representations of engineered features, but their behavior under structural compression and their applicability to regression tasks remain underexplored. In this work, we propose entropy-punctured Bloom Filters, a memory-aware encoding strategy that removes low-variability bit positions identified using empirical entropy. Starting from fixed-length BF encodings of quantized features, the proposed approach produces reduced representations that preserve predictive structure while improving predictive efficiency relative to encoded representation size. We evaluate the approach on diverse regression datasets, comparing raw features, Principal Component Analysis (PCA), Random Projection (RP), and Bloom Filter variants under leakage-free evaluation protocols and approximately matched representation sizes. Performance is assessed using ridge regression, XGBoost, and neural networks, with predictive efficiency measured as R2 relative to encoded representation size per sample. Results show that Bloom Filter encodings remain competitive with classical compressed representations while achieving substantial storage savings. Entropy-based puncturing further reduces representation size with minimal loss in predictive fidelity, yielding improved predictive efficiency. These findings demonstrate that entropy-punctured Bloom Filters provide an effective representation-level compression approach for memory-constrained machine learning. --- ### Bi-Level Routing and Sparse Spatial Attention based Multi-View BEV 3D Object Detection for Autonomous Driving Authors: Jing Zhang, Jiaqi Liu, Zibo Wang Published: 2026-09-12 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.14185v1 Bird's Eye View (BEV)-based multi-view 3D object detection suffers from challenges of computational complexity, multi-scale feature extraction, and efficiency of dense 2D-to-BEV view transformation. To address these problems, this paper proposes an improved BEV 3D object detection algorithm Sparse-BEVNet. Firstly, a Bi-Level Routing Attention (BRA) mechanism is introduced into the image feature extraction network to reduce the computational burden of the backbone. Second, Cascaded Group Attention (CGA) is employed in the feature fusion module, which enhances deep interaction across features of different hierarchical levels without introducing additional computational overhead. Furthermore, a Sparse Spatial Cross-Attention mechanism is adopted to replace the conventional dense view projection pipeline. Experimental results on the public nuScenes dataset demonstrate that the proposed method achieves a mean Average Precision (mAP) of 45.2% and a nuScenes Detection Score (NDS) of 54.5%, corresponding to 3.6% and 2.8% improvements relative to the baseline model, respectively. --- ### ZAPS: Zero-Cost Active Proxy Search for Neural Architecture Search Authors: Hassan Touayouch, Rabie Najem, Mohammed Benjelloun Published: 2026-09-12 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.14184v1 Neural Architecture Search (NAS) automates network design, but evaluating a single candidate requires training it to convergence, making exhaustive search intractable. Zero-cost proxies estimate architecture quality at initialization in seconds, yet a single proxy is noisy, and combining several does not straightforwardly help: proxies are strongly correlated, so naive aggregation compounds their shared errors instead of averaging them out. Existing methods exploit either proxy signals or architectural topology - never both within a single active-learning framework. We introduce ZAPS (Zero-cost Active Proxy Search), a four-stage pipeline that closes this gap. ZAPS (i) selects a compact, non-redundant proxy subset offline via ProxyFit, a greedy anti-redundancy criterion; (ii) seeds the search with a hybrid K-means strategy that balances exploitation and exploration; (iii) re-selects proxies at every iteration by a bootstrapped vote as the labeled set grows; and (iv) ranks candidates with an XGBoost ensemble trained jointly on proxy ranks and one-hot topological encodings, queried through an Upper Confidence Bound (UCB) acquisition function. On NAS-Bench-201 under a budget of B=200 evaluations, ZAPS recovers 52.3% of the true top-100 architectures on CIFAR-10 and 65.8% on CIFAR-100, ahead of every baseline we consider - Random Search, Local Search, REA, BANANAS and TPE - and, on CIFAR-10, with less than half the run-to-run standard deviation of the strongest of them. The advantage is largest where evaluations are scarce: on NAS-Bench-201 it narrows as the budget grows, whereas on the harder NAS-Bench-101, which no method comes close to saturating, it widens instead. All methods are scored by a single criterion: how much of the true top-100 lies among the architectures they actually evaluated. --- ### A Machine Learning Framework for Fault Detection, Isolation, and Severity Prediction of Autonomous VTOL Aircraft Authors: Ripon C. Sarker, Pedram H. Dabaghian, Raman Goyal, Atanu Halder Published: 2026-09-12 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.14180v1 Fault detection in autonomous VTOL aircraft is critical because even minor component degradations can rapidly destabilize multirotor vehicles operating in complex, safety-critical environments, motivating robust fault detection and estimation strategies capable of identifying early signs of rotor damage; however, real-flight fault detection remains challenging due to sensor noise, environmental disturbances, and the nonlinear aerodynamics of multirotor platforms. This study proposes a comprehensive machine-learning framework for rotor fault detection, isolation, and severity prediction using real flight data. A convolutional neural network (CNN) architecture is developed to learn spatio-temporal patterns from multivariate flight dynamics, enabling direct inference of both the faulted rotor and its damage level. The framework is first validated using simulated data generated by a data-generative model, and experimental validation is then performed on a hexacopter by introducing controlled blade-tip breakage. The trained model achieves rotor-wise fault classification accuracies above 99% and severity estimation accuracy of 96% within a 1% tolerance in experimental data, demonstrating strong generalization and supporting real-time health monitoring for autonomous VTOL systems. --- ### A Multi-Stage Agentic Framework for Effective Counter-Narrative Generation and Refinement Authors: Carmel Kronfeld, Sharva Gogawale, Tetsuro Kobayashi, Irad Ben-Gal Published: 2026-09-12 Categories: cs.CL, cs.CY, cs.SI arXiv: https://arxiv.org/abs/2609.14178v1 The rapid diffusion of hate speech and misinformation on social networks challenges democratic societies, since direct suppression efforts may deepen polarization, fuel public distrusts, and strengthen extremist narratives. LLM-driven counter-narratives (CNs) offer a promising way to reduce those risks, yet their effectiveness depends on rhetorical and stylistic choices that remain poorly understood. We present a multi-stage agent-based framework for generating, refining, and evaluating CNs, applied to pro-Russian hate and misinformation narratives on the war with Ukraine and adaptable to other domains. A pilot experiment with human evaluators identifies effective technique style pairings, such as repetition with emotional framing enhancing persuasiveness. Building on these insights, we introduce a multi-agent refinement process that iteratively improves CNs for persuasiveness, emotional engagement, and shareability. After human validation confirmed improvement, an automated safety analysis shows that our refined CNs match or improve on expert-written counterspeech. A simulated experiment then shows that they reduce the perceived strength of pro-Russian narratives and consistently outperform a vanilla LLM baseline, highlighting a pathway toward scalable, narrative-specific interventions against hate speech and misinformation. Code and data accompanying this work are publicly available at https://github.com/carmelkron/inlg2026-counter-narratives. --- ### Inherited Heads: Audio language models track speakers with their text backbone's attention, and an attention-mass ranking retrieves a different set Authors: Bojro Das Published: 2026-09-12 Categories: cs.LG, cs.CL, cs.SD arXiv: https://arxiv.org/abs/2609.14174v1 Asked to describe what one of six speakers in a recording talks about, audio language models describe the right one on 6 to 16% of trials, below the 16.7% a guess would give. Adding a fixed bias to the attention logits of a hundred heads, under a tenth of the model's and with no training, redirects the description to whichever speaker we choose, on 90.7% to 99.0% of trials. Those heads are largely not specific to audio. Rank the text-only language model an audio model was built from, or a released model of the same family, on a written version of the task, take its top hundred heads, and carry them over unchanged: they redirect the audio model on 80.8% to 95.0% of trials, with nothing about audio entering the selection. The audio and text head sets share 66 to 74 of 100 where chance would give about 20, and the shared part alone reproduces almost all of the steering. What that does not show is that sharing is what makes the heads work: an equal-sized draw from the same discovered hundred does nearly as well, and none of our three models separates the two explanations. A second finding concerns how such heads are found. Ranking heads by how much attention they place on the segment asked about, as an established score does, or by how much of their attention moves with the question, as a per-head normalised variant does, gives top hundreds that share 69, 37 and 4 heads across our three models. In Ultravox, where they share 4, the established score's heads leave output the judge cannot place on any segment on 69.7% of trials, against 40.0% with no intervention and 1.0% for the normalised variant. That is one arm of six; on the other five the established score steers above a random draw. --- ### CyFM: Cylindrical Optimal Transport for Few-Step Complex-Valued Flow Matching Authors: Marcel Musiałek, Iga Wolanin, Damian Ryczko, Anna Grelewska, Oleksii Furman Published: 2026-09-12 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.14171v2 Complex-valued signals like MRI and audio spectrograms are typically modelled as flat two-channel Euclidean data. The inherited Euclidean metric $dA^2 + A^2 dθ^2$ vanishes at the origin, leaving phase unpenalised exactly where the signal is weakest. We replace it with the decoupled product metric $dA^2 + dθ^2$ on the cylindrical closure $(0, \infty) \times S^1$, which stays non-degenerate at $A = 0$. We measure what this substitution costs and buys. Exact analytical bridges across synthetic fields, fastMRI knee data, and LibriSpeech spectrograms show Cartesian paths induce a heavy-tailed angular velocity distribution (Pareto index $\approx 1$). Under independent coupling, 43%-49% of signal energy falls on paths turning faster than $π$ rad per unit time. Cylindrical paths never reach this speed. We formulate Cylindrical Flow Matching (CyFM) to strictly bound the angular regression target, coupling noise and data via exact minibatch Optimal Transport jointly over whole fields. This coupling reduces few-step generation error by 3%-60%. CyFM achieves lower generative error than the best Cartesian baseline at every step up to $k = 8$ on synthetic fields and speech spectrograms, with all seeds separated. On knee MRI, the single-step advantage is 1.8x. At convergence ($k = 100$), the two geometries show no significant difference. Finally, a prior-only control exposes the cost of flat parametrisation: on synthetic fields, a single Cartesian Euler step performs worse than the unintegrated noise prior (0.376 vs. 0.150). --- ### On Large-Scale Multiple Testing Over Networks: A Non-Asymptotic Approach Authors: Mehrdad Pournaderi Published: 2026-09-12 Categories: stat.ME, cs.IT, eess.SP, stat.ML arXiv: https://arxiv.org/abs/2609.14170v1 Distributed multiple testing asks $N$ sites to control a global false discovery rate (FDR) under a tight communication budget. The greedy interval-aggregation algorithm of Pournaderi and Xiang (2024) solves this asymptotically but can violate $\mathrm{FDR}\leα$ at finite samples. We trace the violation to a winner's-curse bias in the selected density statistics, of exact order $Θ(m^{-1/4}\sqrt{\log m})$ at the standard bandwidth $\varepsilon\asymp m^{-1/2}$, with $m$ the total number of p-values in the network. Cross-Fit Greedy Aggregation (CFGA) eliminates the curse by selecting the nested rejection family on one half of each node's data and scoring it on the other, achieving finite-sample $\mathrm{FDR}\leα$ when per-node null rates are known; an inflated variant covers the plug-in setting at a vanishing $η=1/m$ slack. BONuS-GA instead masks a bag of synthetic uniform nulls calibrated by counting knockoffs, so every p-value serves both selection and inference; a per-node budgeted variant removes all oracle input, controlling $\mathrm{FDR}\leα$ for any data-independent bag. Aggregating the CFGA folds by e-values over random splits (e-CFGA) removes the Bonferroni factor and averages the split randomness. All variants keep the $O(\sqrt{m}\log m)$ communication budget, up to e-CFGA's $O(\bar{R}\log m)$ reporting round; empirically, BONuS-GA dominates at moderate-to-large per-node samples and CFGA with adaptive bandwidth at small ones. --- ### Towards Evolving Context Parameterization for Large Language Models Authors: Xiaobing Shi, Zherui Li, Yiming Jiang, Kun Wang, Yufei Guo Published: 2026-09-12 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2609.14168v1 Context parameterization enables large language models (LLMs) to internalize contexts into reusable model parameters, avoiding repeated processing across subsequent queries. However, existing methods typically assume static contexts and lack explicit mechanisms for distinguishing validity states under continual updates. To study this real-world scenario, we formalized the Memory Updating with Sequential Evolution (MUSE) task and constructed MUSE-bench to evaluate update incorporation and unaffected-information preservation. The resulting challenge requires preserving the global state while adjusting the contribution of memory evidence. Motivated by this, we proposed PLUME, a training-free method that constructs a global update representation, activates memory evidence to form a local parameter view, and adaptively integrates their predictions during decoding. Comprehensive evaluation on MUSE-bench demonstrated PLUME's effectiveness in sequential evolution settings, yielding relative improvements of 29.9% in average ROUGE-L Recall and 54.9% in LLM-as-a-Judge. Our codes are available at: https://github.com/xiaobingshi-LLM/PLUME. --- ### 3D Gait-Based Autism Classification Using Attention-Enhanced Deep Learning with Cross-Fold Statistical Stability Analysis Authors: Md Nadim Mahamood, Md Arif Shahriar, Md Parvej Sikder, Md Rasul Islam, Md Shafi Ud Doula, Md Ashraful Alam, Kamrul Hasan Published: 2026-09-12 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2609.14159v2 Autism Spectrum Disorder (ASD) is a neurodevelopmental condition whose early diagnosis remains challenging because conventional clinical assessments are often subjective, time-consuming, and require expert evaluation. Gait provides a promising non-invasive behavioral biomarker for auto- mated ASD screening; however, existing studies have primarily relied on single-dataset evaluations, convolutional architectures, and descriptive summaries of cross-validation performance without formally assessing fold-to-fold stability. This study addresses these gaps with an attention-enhanced Transformer framework for ASD classification, evaluated on two structurally different 3D gait feature representations: precomputed statistical gait descriptors and raw biomechanical ground-reaction- force measurements. Under five-fold cross-validation, the proposed framework achieved 99.00% accuracy, 99.02% precision, 99.00% recall, 99.00% F1-score, and 99.00% specificity on the public Kinect-based benchmark, exceeding the performance of the compared state-of-the-art methods. On the independent private force-plate dataset, it achieved mean values of 95.00% accuracy, 93.81% precision, 96.67% recall, 95.13% F1-score, and 93.33% specificity. --- ### When Tools Get in the Way: The Effect of Unnecessary Tool Availability on LLM Answering Authors: Saanvi Paturi, Arsen Kenzhebayev, Arham Sethi, Vyas Raina, Ivaxi Sheth, Vatsal Raina Published: 2026-09-12 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.14157v1 Large language models (LLMs) are increasingly deployed with external tools that extend what they can do beyond their own knowledge. Tools help on tasks that need external information, but their availability may also change how a model handles questions that do not need them. Prior work has mostly asked whether models select and use tools appropriately; whether an unnecessary tool changes the correctness of answers has received less attention. We ask whether making a related but unnecessary tool available affects a model's ability to answer from its own knowledge, and whether a preceding tool interaction changes this behaviour. We construct 500 query pairs across 10 knowledge domains. Each pair consists of a tool query, which needs the domain's tool, and a closed-domain query, which does not. Six LLMs are evaluated with the tool unavailable, available, and available after a prior tool call. Across 3,000 baseline trials the pooled answer rate is 98.2%. When an unnecessary tool is available it falls to 63.5%, with large differences between models. The decrease occurs even when the tool is rarely called, so it cannot be explained by unnecessary tool invocation alone. A one-sentence scope-aware system instruction recovers most of the lost answers. --- ### Signatures of Steerability in Activation Space of Language Models Authors: Prajjwal Bhattarai, Tuka Alhanai Published: 2026-09-12 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.14151v1 Steering language models using a set of contrastive representations has been a canonical and computationally efficient method for controlling model behavior. Despite this success in controlling certain model behaviors, the effectiveness of activation steering varies markedly across concepts; the generalization properties of steering vectors are often considered a function of the dataset used to construct them. We make this dataset-dependence claim more rigorous and show that simple separation metrics strongly correlate with the downstream steerability of language models across diverse settings, even after controlling for layers and dataset effects. Beyond prediction, we provide evidence from a synthetic superposition experiment that separation metrics are strongly correlated with alignment between the empirical and true feature direction. Our results suggest that simple separability statistics can serve as practical diagnostics for when steering vectors are likely to work. --- ### When Faster VLA Deployment Changes Closed-Loop Behavior: Task Success-Latency Analysis of SmolVLA Across PyTorch and ONNX Variants Authors: Rafiqul Islam Published: 2026-09-12 Categories: cs.RO, cs.LG arXiv: https://arxiv.org/abs/2609.14146v2 Vision-language-action (VLA) deployment can reduce inference latency while changing closed-loop task behavior. We evaluate HuggingFaceVLA/smolvla_libero on an RTX 2060 (6 GB) in LIBERO Spatial and Object (MuJoCo 3.3.2, LeRobot 0.6.1, seed 42), comparing PyTorch+AMP with ONNX Runtime CUDA Execution Provider (CUDA EP). The main evaluation uses 100 episodes/suite; a paired rollout uses 300 episodes/suite. PyTorch+AMP reaches 70.0%/88.0% Spatial/Object success at 1181 ms p99. Requested-FP16 and requested-INT8 ONNX reduce tether-inspect p99 to 601 ms and 532 ms, while Spatial success falls to 41.0% and 40.0% and Object remains at 89.0%. A graph audit shows those artifacts are byte-identical FP32 graphs, so the requested-INT8 row is not operator-level INT8 quantization. A static language-width ablation (16/24/32 tokens) yields Spatial success of 41.0%, 75.0%, and 71.0%; widths 24 and 32 recover much of the Spatial drop while Object success and uniform-bench latency stay approximately stable. Width-24 ONNX Spatial success is comparable to the PyTorch+AMP baseline at roughly half the latency (Wilson intervals overlap; two-proportion chi-squared p=0.53). Context width is an important contributor in this stack; it does not account for every PyTorch-vs-ONNX difference. Deployment evaluation should jointly report latency, artifact inspection, interface constraints, and closed-loop success. Code: https://github.com/rafiqul713/smolvla-libero-onnx. --- ### A New Transformer-Based Approach for Audio-Based Kinship Verification and a New Uncontrolled Mandarin Kinship Speech Dataset Authors: Qiyang Sun, Langqing Zhang, Yupei Li, Björn Schuller Published: 2026-09-12 Categories: cs.SD, cs.AI, cs.LG, eess.AS arXiv: https://arxiv.org/abs/2609.14145v2 Kinship verification is a task involving determining whether two individuals share a first-order kin relation. To tackle this task, we propose CONVTRAP-TN, a new architecture for audio-based kinship verification, and conduct an ablation study on the proposed model. To the best of our knowledge, we are the first to apply the successful transformer architecture to the task of audio-based kinship verification. Furthermore, we also collect a custom speech dataset, ARKIN, which accurately reflects everyday recording conditions. We do this because only a few speech datasets with kinship labels currently exist, all of which either source extremely noisy in-the-wild data from the internet, or instruct speakers to record in specific environments. These settings fail to reflect real-world scenarios where users record on personal devices under unrestrained conditions. Additionally, we perform a series of preliminary baseline experiments on the collected dataset, including speaker verification and recognition, speech recognition, age estimation, and kinship verification, as well as cross-dataset kinship verification experiments to show that existing methods are not robust across datasets. --- ### One Size Does Not Fit All: Setting Inference Depth from the Questions a Deployment Actually Asks Authors: Jerry Kaplan Published: 2026-09-12 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2609.14144v1 A transformer language model is trained to respond to any prompt, but each deployment asks only a narrow range of questions: a support assistant sees delivery complaints, a coding tool sees Python. Every deployment nonetheless pays the same computation per token. This paper measures how much of that cost is avoidable when the range of prompts is known in advance. The mechanism examined is early exit: a small, trained component - called a readout - is attached to an intermediate layer and proposes a token, and a confidence test decides whether to emit it or to run the remaining layers. The models are frozen, and the only supervision used is the model's own output on ordinary traffic. Three findings are reported. First, achievable savings depend strongly on the kind of traffic: at half depth on a 1.5-billion-parameter model, 96 percent of tokens could be emitted early for arithmetic word problems and 8 percent for Chinese-language explanations, at matched token-level fidelity to the full model (a measure whose limits the third finding exposes). Second, of three ways a deployment might use knowledge of its traffic, only customizing the threshold for exiting early is worthwhile: calibrating it per deployment raised exit rates by up to 59 percentage points across three models, and by more than 10 points on most corpora tested. Third, token-level fidelity - the standard evaluation measure in the early-exit literature - fails in domains where tokens can be checked against ground truth: on arithmetic word problems, three models each answered sixty questions correctly when run in full, and between 10 and 28 correctly under early exit, in the configuration that scored highest on fidelity. The intended setting is small models on personal devices, where generation is limited by memory bandwidth rather than computation. --- ### T-SMART: Mechanism-Level Attribution for Tool-Augmented Time-Series Question Answering Authors: Ivan Delgado, Himansi Gupta, Bishal Khatri, Niharika Sapre, Lameta Shamoon, Onat Gungor, Tajana Rosing Published: 2026-09-12 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.14142v1 Large language models (LLMs) can struggle with time-series question answering (TS-QA), especially when numerical signals are serialized as text and require explicit computation. Tool-augmented approaches improve performance, but existing systems often intertwine language reasoning, computation, and perception, making it difficult to determine which components drive the gains. We present T-SMART, a neurosymbolic framework that separates these roles: a frozen LLM interprets questions and selects operations, deterministic tools perform numerical computation, and structured perception is invoked only when needed. Controlled paired ablations show that deterministic computation provides the dominant benefit, improving accuracy by 31.7 percentage points over direct LLM reasoning on serialized time series, while language understanding and perception offer smaller complementary gains. These results indicate that tool-augmented TS-QA benefits primarily from reliable numerical execution rather than additional language-model reasoning and provide a controlled framework for analyzing component contributions in neurosymbolic time-series systems. --- ### Riemannian ascent--descent for nonconvex nonconcave minimax landscapes: convergence to basin saddle points and applications to distributionally robust optimization Authors: Rishabh Dixit, Pranav Upadrashta, Alex Cloninger Published: 2026-09-12 Categories: math.OC, cs.LG, stat.ML arXiv: https://arxiv.org/abs/2609.14141v1 We study a class of distributionally robust optimization (DRO) problems for the statistical risk problem, formulated as minimax problems over the product of a Euclidean space and a Riemannian manifold. Because the resulting minimax landscape is nonconvex nonconcave in general, no globally convergent first order method is known to be available. We instead introduce the notion of a \emph{basin saddle point}, a Nash equilibrium defined locally on the Cartesian product of a $δ$ basin around a connected component of the local minima critical set and a geodesic ball on the measure manifold. We develop an abstract convergence framework for a Riemannian gradient ascent multistep descent iteration to a basin saddle point under a local Łojasiewicz type growth condition, with exponent $β\in (1,2)$, in the $δ$ basin around connected components of the local minima critical sets. Under Lipschitz regularity of critical sets we establish linear convergence for $β= 2$ and polynomial convergence for $β\in (1,2)$ to a basin saddle point, with explicit dependence on the sectional curvature of the manifold. We then instantiate this framework for the statistical risk DRO problem over Gaussian measures, where the ambiguity set is naturally modeled as the product of Euclidean space and the Bures Wasserstein manifold of covariance matrices, which we relax to a penalized DRO formulation. We derive nonasymptotic Hessian estimates for the resulting Lagrangian, establish existence and local uniqueness of its maximizer, and prove that an alternating Riemannian gradient scheme converges to a basin saddle point of the penalized DRO problem, recovering the linear and polynomial rates of the abstract theory with all constants explicit in terms of data dimension, loss moments, and the reference covariance. --- ### LIMBO: Lifelong Inference-Time Memory and Budget Optimization for LLM Agents Authors: Siddharth Sharma, Nilesh Prasad Pandey, Onat Gungor, Tajana Rosing Published: 2026-09-12 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.14138v1 As LLM agents become integrated into increasingly complex workflows, they must continually acquire new capabilities while retaining competence on previously learned tasks. Lifelong agents address this through experience replay, injecting past interactions into the prompt to leverage prior experience during inference. However, replay is not free: every replayed trajectory competes with retrieval, reasoning, tool use, and verification for the same limited prompt and compute budget, making effective resource allocation essential. Existing approaches allocate these resources using fixed replay policies, regardless of whether replay is beneficial for the current task. We identify this as inference-time memory allocation, a distinct problem class for lifelong agents, and introduce LIMBO: the first online framework to our knowledge that treats memory as a controllable inference-time resource and jointly optimizes memory strategy and inference budget for each incoming task. Unlike prior approaches that fix the replay policy or require model weights, teacher supervision, or offline retraining, LIMBO learns this allocation online in a single pass, explicitly balancing task performance and inference cost without modifying the underlying agent. Across three LLM backbones on LifelongAgentBench, LIMBO achieves better cost-accuracy tradeoffs than state-of-the-art memory-augmented baselines and nearly matches all strongest such baselines at up to ~83% lower inference cost (~53% on average). LIMBO adapts its policy across models and environments without retraining, demonstrating that effective allocation can be learned online rather than manually specified. --- ### Deformable 2D Gaussian Splatting for Efficient 4K Video Compression Authors: Chenhao Zhang, Fengqing Zhu Published: 2026-09-12 Categories: eess.IV, cs.CV arXiv: https://arxiv.org/abs/2609.14129v1 Ultra-High-Definition (UHD) video presents significant challenges for efficient storage and real-time decoding. Learning-based methods, such as Neural Video Compression (NVC) and Implicit Neural Representations (INR), achieve competitive rate-distortion performance but suffer from high decoding latency and excessive memory usage. Meanwhile, Gaussian Splatting has recently attracted attention in the computer graphics community due to its ultra-fast rendering and high-fidelity visual quality. Despite these advantages, its application in video compression remains largely unexplored. To bridge this gap, we propose a real-time video compression framework that represents and compresses a Group of Pictures (GOP) using a coarse-to-fine multi-scale 2D Gaussian Splatting (2DGS) structure coupled with a lightweight deformation network. Experiments demonstrate that our method delivers rate-distortion performance in LPIPS that surpasses H.265 and other state-of-the-art learning-based video compression methods. Our work demonstrates the potential of Gaussian Splatting as a practical solution for efficient high-resolution video compression. --- ### Exact Finite Attention Responses From RoPE Derivatives Authors: Julie Huang, Maggie Chlon, Gregory Gutin, Leon Chlon Published: 2026-09-12 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2609.14127v1 We derive exact local responses for attention interventions, allowing candidate edits to be scored from a cached baseline and one backward pass. The starting point is the RoPE derivative $\partial_p z(p) = A z(p)$: its integral gives the finite positional displacement, which we carry through the softmax without linearising either rotation or normalisation. The resulting predictions achieve 95.36--96.52% sign accuracy across 92,160 executed positional edits on 768 held-out prompt sets, reducing answer-margin MAE by 73.6--82.5% against the positional Jacobian and by 36.2--50.9% against zero. For simultaneous key and value edits, the same divided-difference calculus isolates the interaction term $C_{KV} = \sum_j (p'_j - p_j)\,\varepsilon_j$, which is omitted by adding separate attributions. Retaining it reduces downstream margin MAE by more than a factor of nine in every setting of a 5,120-intervention sweep across two Qwen sizes, two tasks, and multiple layers; reductions against a quadratic interaction correction are 75.9--98.5%. Exactness concerns the edited attention write; downstream predictions contract that response with a baseline gradient and are evaluated by native execution. The calculus also yields a KL certificate for local approximation error, an exact query-conditioned gradient-step representation whose curvature identifies attention-preserving query directions, and minimum-norm query control. Sparse evaluation supports candidate ranking and cache decisions under explicit local distortion criteria. --- ### To do($x$) or not to do($x$): Medical Image Counterfactuals for Dataset Augmentation Authors: Yasin Ibrahim, Robin J. Evans, Konstantinos Kamnitsas Published: 2026-09-12 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.14124v1 Medical image analysis is often hindered by biased datasets, which can lead to biased models and limited clinical applicability. A promising strategy for mitigating such biases is to augment training data with synthetic images. Counterfactual (CF) generation is one such strategy, though the term is used in two different senses: in some works, CFs are produced through causality-based interventions derived from structural causal models, whereas in others, they are produced by non-causal image edits or conventional conditional generative models, such as altering anatomy or adding pathologies. In this work, we study this distinction and evaluate its practical consequences for medical image augmentation. We compare three conditioning strategies: $\textit{Deterministic}$, which changes selected variables while holding the remaining variables fixed; $\textit{Undirected}$, which updates variables according to learned statistical associations without assigning causal directions; and $\textit{Causal}$, which propagates interventions along a directed causal graph. We analyse how these choices affect the resulting images, and explore when causally grounded methods improve dataset augmentation or bring limited benefit. In particular, we assess downstream performance and fairness, where fairness refers to reduced sensitivity to dataset biases across sensitive subgroups. Our experiments demonstrate that using a causal approach to synthetic training data generation can lead to tangible benefits, with these insights offering valuable guidance to machine learning practitioners for the effective design of data generation protocols. --- ### SignMimic: Robust High-Quality Sign Language Motion Generation via Human-Shape-Oblivious Pose Transfer Guidance Authors: Zhewen He, Junyi Yu, Haomian Huang, Zhenhua Li, Yi Fang Published: 2026-09-12 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.14122v2 We study the challenge of sign language video mimicking: given a driving video and a single reference frame, synthesize a video where the target signer reproduces the source motion while preserving identity and linguistic form. Prior pipelines entangle rigid motion, non-rigid deformation, and view-dependent completion in a monolithic generator, causing handshape drift and spatio-temporal instability. We present SignMimic, which (i) applies a TNet-based model to study SE(3) rigid canonicalization to stabilize global pose, (ii) performs non-rigid adaptation in a canonical space to preserve fine-grained articulators (hands/face) and coarticulation via NIF2D, and (iii) uses Pose-MAE-style completion before conditional video diffusion. This factorization injects geometric and linguistic priors, yielding shape and spatio-temporal consistency. On several large-scale datasets (ASL 50K, How2Sign, CSL News), SignMimic achieves state-of-the-art-level performance on video quality, identity similarity, and frame continuity while also achieving minimal loss when performing back translation (SLT) on generated videos. Ablations confirm the role of rigid canonicalization, non-rigid adaptation, and completion. Code is available at https://anonymous.4open.science/r/UniSignMimicTurbo-6088; model checkpoints and video examples will be released. --- ### Semantic Knowledge Technologies: what the Semantic Web lost sight of, and what it never had Authors: Achille Zappa Published: 2026-09-12 Categories: cs.AI, cs.DB, cs.IR arXiv: https://arxiv.org/abs/2609.14121v1 The Semantic Web set out to give information a machine-interpretable form so that software could integrate and reason over it. Its standards became scientific knowledge infrastructure, but the machine competence it promised did not follow, and the systems now answering questions over scientific knowledge are language models holding no inspectable account of what they know. This paper argues the original goal was right and the technical programme incomplete, states what is missing, and names the extended programme Semantic Knowledge Technologies: the same technical core carried out of its web-publishing origin and applied to knowledge wherever held. The diagnosis is that the standards formalised truth while omitting three things: the conditions under which a claim holds, the operations its terms permit, and any account of what a base covers. Without conditions, contradiction and applicability cannot be judged; without operational grounding, holding a statement confers no ability; without declared coverage, a system cannot recognise the boundary of its own content, which under the open-world assumption cannot be inferred. The paper fixes the word understanding to five measurable tests (check, connect, derive, act, delimit) and sets out a seven-layer architecture in which the first three layers are enabling and the rest the cognitive capabilities they make possible. It then defines three terms the programme implies: Large Knowledge Model, a model whose unit of output is a reference to an addressable claim, not a token; SLKM, the knowledge base an agent builds for itself from declared sources; and Semantic Artificial General Intelligence, stated as a falsifiable position about necessary conditions, not a system. A graded ladder replaces the untestable word general. It is offered as a research agenda, with its weakest points and refutation condition named. --- ### Talking to Me or Someone Else? Rethinking Talk-to-Me Detection in Egocentric Videos Authors: Feiyu Du, Xi He, Jia Li, Yapeng Tian, Weili Wu Published: 2026-09-12 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.14118v1 Online understanding of who is talking to the camera wearer is a key capability for egocentric social interaction. However, existing talk-to-me (TTM) studies are commonly formulated as offline clip-level recognition, which is poorly aligned with online interaction and overlooks the diverse non-TTM speaking states that naturally arise in egocentric videos. In this paper, we revisit this problem by reformulating it as an online, frame-level prediction task. Instead of treating TTM as a binary problem against a single negative class, we model it in the presence of diverse and previously underexplored non-TTM states, such as talking-to-others, self-talking, and background conditions. To support this new formulation, we construct an Online TTM Dataset consisting of 406 egocentric video clips with approximately 900K annotated frames, each labeled with frame-level social interaction categories (e.g., background, TTM, talking-to-others, self-talking), by extending the Ego4D social interaction benchmark. In this benchmark, we evaluate five adapted baselines and develop a new model that integrates social cues across modalities. Experimental results show that our multimodal model, which jointly leverages audio, visual, and speech-semantic cues, achieves 75.5% frame-level F1 on TTM, outperforming strong baselines and enabling a systematic analysis of how different speaking states affect TTM recognition. --- ### Real-Time Synthesis of Robust Controlled Invariant Sets for Monotone Systems Authors: Yasin Sonmez, Mahmoud Khaled, Majid Zamani, Murat Arcak Published: 2026-09-12 Categories: eess.SY, cs.AI, cs.RO arXiv: https://arxiv.org/abs/2609.14115v1 Safety-critical control of autonomous systems requires formal safety certificates, such as controlled invariant sets, that must be computed online as conditions change. Although standard synthesis algorithms scale poorly with state dimension, monotone dynamical systems with lower-closed safety specifications allow for accelerated computation of controlled invariant sets. In particular, lazy fixed-point algorithms exploit monotonicity and track only the antichain basis of the set. However, membership tests and redundancy checks against an evolving basis remain major bottlenecks. We introduce a threshold-function reformulation in which a lower-closed set on a d-dimensional grid is represented by its column heights along a designated axis. This reformulates the greatest-fixed-point iteration as independent one-dimensional binary searches, one per grid column, yielding an embarrassingly parallel iteration with asymptotically lower computational complexity than the lazy fixed-point algorithm. Experiments synthesize invariant sets on 3D grids with 10^9 cells in under 50 ms and 10^14 cells in under two minutes. We further demonstrate online re-synthesis in a safety-informed model predictive controller example. --- ### Parameter-Efficient Fine-Tuning of Foundation Models for Liver Tumor Segmentation in CT Authors: Ramtin Mojtahedi, Mohammad Hamghalam, Jacob J. Peoples, Richard K. G. Do, Amber L. Simpson Published: 2026-09-12 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.14106v1 We evaluated parameter-efficient fine-tuning (PEFT) of the Segment Anything Model (SAM) for liver tumor segmentation in abdominal CT of colorectal liver metastases. We compared Low-Rank Adaptation (LoRA), 4-bit Quantized LoRA (QLoRA), a convolutional adapter (Conv-Adapter), and our Directional Spectral Top-K adapter (DiSCo), training only adapters while freezing the SAM backbone. DiSCo derives spectral bases from singular value decomposition of row-normalized weights and learns rank-gated spectral coefficients, per-output magnitude offsets, and a spectral gain, with optional Top-K rank selection at inference and 0.14 M trainable parameters. We benchmarked five prompting regimes: no prompt, single-point, multi-point, and bounding boxes at intersection over union 0.50 and 0.75. Conv-Adapter and LoRA achieved the highest accuracy (overall Dice 0.793 and 0.792; single-point Dice 0.795 and 0.792; 95th-percentile Hausdorff distance (HD95) 32 mm). QLoRA was close (overall Dice 0.766; single-point Dice 0.768; HD95 36.41 mm), with 0.91 M trainable parameters, 120 ms latency, and 4.9 GB peak memory. DiSCo achieved the highest Dice per million trainable parameters (4.66), with overall Dice 0.653, single-point Dice 0.698, and HD95 49.53 mm. These results show an accuracy-efficiency trade-off and support PEFT for liver tumor segmentation with reduced adaptation costs when compute and labeled data are limited. Code: https://github.com/Ramtin-Mojtahedi/PEFT-SAM-Liver-CT --- ### Accelerating HKTex without Mesh Eigensystems: Local Unfolding and Randomized Thermal Features Authors: Zhewen He, Junyi Hu, Yi Fang Published: 2026-09-12 Categories: cs.CV, cs.GR arXiv: https://arxiv.org/abs/2609.14105v1 Heat Kernel Textures (HKTex) represent surface appearance with intrinsic anisotropic kernels, but evaluate them using 50 global Laplace-Beltrami eigendecompositions and a resident basis of shape (50,V,256). We study two complementary ways to remove this bottleneck while leaving the trainer, GeodesicOpt, density control, and compositing unchanged. LocalHK exploits the measured locality of trained kernels and replaces spectral evaluation by radius-bounded hinge unfolding and an analytic log-map kernel. On 10 Objaverse meshes and an 8-mesh low-poly holdout, it changes mean view PSNR from 31.35 to 32.00 and from 29.82 to 30.76 dB, respectively, while reducing initialization by 40.5 times and enabling a 749,570-vertex proxy-backed run where the spectral baseline fails. ThermalRF instead preserves the discrete anisotropic heat semigroup: GPU sparse Chebyshev actions and randomized range finding construct global low-rank heat factors without mesh-sized eigenvectors, and a compiled evaluator mixes four neighboring thermal responses. On spot and a thin-stem challenge, ThermalRF reduces end-to-end preprocessing, initialization, and 5,000-step optimization by 29.3% and 24.4%, with every surface, atlas, or view PSNR change within 0.12 dB and training allocation reduced by about 90%. The two routes expose a useful design choice: maximal locality and scale versus fidelity to the thermal PDE. Broader thermal-feature evaluation and real large scenes remain future work. --- ### A Voxel-Spacing-Aware Extension of PyRadiomics for Anisotropic Texture Analysis Authors: David Corral Fontecha, Juan Miranda Bautista, Pablo Menendez Fernández-Miranda, Andrea Trapote Fernandez, Lara Lloret Iglesias, Jose A. Vega Published: 2026-09-12 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.14103v1 Radiomic texture features are commonly extracted from anisotropic CT and MRI acquisitions, where identical voxel offsets may represent different physical distances. We implemented and validated a voxel-spacing-aware extension of PyRadiomics that incorporates spacing information without generating interpolated gray levels. The framework operates across the Python frontend, C wrapper, and computational backend. GLCM uses anisotropy-relative feature-level angular aggregation, NGTDM uses anisotropy-relative weighted neighborhood averaging, and GLRLM, GLDM, and GLSZM are computed on a finite-volume zero-order-hold representation derived from the native anisotropic grid. Synthetic 3D phantoms were used for software validation. The modified implementation reproduced standard PyRadiomics exactly when spacing-aware mode was disabled and remained equivalent under isotropic spacing across 75 texture features. Under anisotropic spacing, the method selectively modified texture families and was numerically distinct from nearest-neighbor, linear, and B-spline resampling. Computational profiling showed moderate runtime and memory increases, while sensitivity analyses quantified finite-volume rounding effects and confirmed that spacing-aware differences persisted across binWidth settings. The framework provides a backward-compatible technical basis for future evaluation of spacing-aware radiomics in heterogeneous medical imaging datasets. --- ### RA-CoA: Training-free Fashion Image Captioning via Retrieval-Augmented Chain-of-Attributes Authors: Abhirama Subramanyam Penamakuri, Shreya Shukla, Anand Mishra Published: 2026-09-12 Categories: cs.CV, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.14100v1 Fashion Image Captioning (FIC) plays a vital role in enhancing user experience and product search in e-commerce platforms. Unlike natural scene image captioning, FIC requires fine-grained visual reasoning and knowledge of domain-specific terminology to capture subtle attributes such as neckline and closure types, graphic patterns, and dress silhouettes. Moreover, as fashion inventories evolve rapidly with new trends, styles, and frequently emerging vocabulary, developing training-free captioning solution becomes essential for scalability and real-world adaptability. Instruction-tuned vision-language models (VLMs) offer a promising solution to fashion image captioning dueto their strong zero-shot capabilities and natural language fluency. However, these general-purpose models often lack attribute-level coverage and precision, and tend to hallucinate or misidentify fine-grained fashion details, making them less suitable for high-fidelity applications like product cataloging or personalized recommendations. To address this, we propose RA-CoA (Retrieval-Augmented Chain-of-Attributes), a novel, training-free framework that disentangles fashion image captioning into two interpretable stages: (i) retrieval of relevant attribute sets from a product knowledge base, and (ii) attribute-level reasoning to generate the final caption. RA-CoA is a model-agnostic approach that works with frozen VLMs to improve fine-grained attribute precision in product captions without the need for fine-tuning. Extensive evaluations across diverse VLM model families under different prompting paradigms demonstrate that RA-CoA significantly improves caption quality, achieving an average gain of 26.3% METEOR score over zero-shot captioning. We make our code publicly available. --- ### Neuron Activation-based Computation of Logical Explanations for Deep Neural Networks Authors: Tomáš Kolárik, Faezeh Labbaf, Fabrizio Leopardi, Grigory Fedyukovich, Michael Wand, Natasha Sharygina Published: 2026-09-12 Categories: cs.LO, cs.LG arXiv: https://arxiv.org/abs/2609.14099v1 Formal explainability of classifying neural networks (NNs) is an active area of research, providing explanations with provable guarantees of the classification within continuous regions of the input feature space. However, the existing techniques are either limited to individual input features without guarantees on their relations or the provided solutions fail to scale to deep architectures. This paper addresses these issues by introducing a flexible symbolic framework for an efficient, guided computation of explanations of the NN behavior, parametrized by the activations of internal neurons, and using logical engines such as SMT solvers. Unlike prior methods that rely on specialized NN verifiers, our method yields explanations that are not restricted in shape. Our algorithm is implementable on top of a general-purpose logical solver, isolating the NN-specific encoding from the algorithmic framework. We experimented with a wide range of benchmarks from the domains of image recognition and medicine, illustrating the advantages of the new method, particularly in computational efficiency. Notably, our approach enables logical explanation of deep networks not amenable to prior logic-based methods. --- ### Bridging the Synthetic-to-Real Gap for Few-Shot Cryo-ET Classification Authors: Siddhant Bharadwaj, Ashish Vashist, Rashi Singh, Pranav Vinodh, Nishanth Artham, Runmin Jiang, Xingjian Li, Min Xu Published: 2026-09-12 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2609.14097v1 Subtomogram classification in cryo-electron tomography (cryo-ET) is a challenging problem due to the scarcity of labeled examples. While cryo-ET simulators can be adopted to generate unlimited synthetic data, the substantial domain gap between synthetic and real subtomograms hinders its practical utilization. In this work, we propose a novel synthetic-to-real adaptation framework with a learnable transformation module, bridging this gap at both the input and feature levels. Extensive experiments demonstrate that our method consistently outperforms existing transfer learning baselines in few-shot settings. --- ### NeuroFlex: Lossless Element-Level ANN-SNN Co-Execution for Efficient Sparse Inference Authors: Varun Manjunath, Pranav Ramesh, Gopalakrishnan Srinivasan Published: 2026-09-12 Categories: cs.AR, cs.LG arXiv: https://arxiv.org/abs/2609.14092v2 Sparse DNN accelerators specialize in ANN or SNN execution, leaving energy or latency on the table when workload characteristics vary within a layer. Hybrid accelerator designs that switch modes at layer or tile granularity suffer from low PE utilization since one core type idles whenever the other is active. NeuroFlex is the first accelerator to assign every output element independently to ANN or SNN execution mode with zero accuracy loss. We extend integer-exact ANN-SNN equivalence from layers to individual output elements, thereby enabling mode switching with no conversion error. An offline cost-guided scheduler scores each element by its marginal energy-delay trade-off and packs work across PEs, achieving 97-99% PE utilization compared to 40-45% for layer-wise hybrids. NeuroFlex reduces EDP by 57-67% over a strong ANN-only baseline and delivers up to 2.5x speedup over a dual-sparse SNN-only baseline. Our cost-guided scheduler improves throughput by 16-19% over random element assignment across vision, language, and transformer workloads. --- ### Just add noise: Debiasing tree-based variable importance in mixed data Authors: Jiahe Li, Omar Melikechi Published: 2026-09-12 Categories: stat.ML, cs.LG, stat.ME arXiv: https://arxiv.org/abs/2609.14083v1 Variable importance scores from tree-based methods such as random forests favor continuous predictors over categorical ones. We present a theoretical analysis of this bias and propose a simple remedy: add a small amount of noise to each categorical predictor. The correction is demonstrated on a variety of simulated and real-world datasets and combined with integrated path stability selection to perform variable selection with false discovery control for mixed data. --- ### Adapting Open-Weight MLLMs to Generate Point Prompts for Electron Microscopy Segmentation Authors: Samia Mohinta, Albert Cardona Published: 2026-09-12 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.14080v1 Promptable models such as microSAM segment electron microscopy (EM) images from point prompts, but automation requires generating prompts without user input. We ask whether open-weight multimodal large language models (MLLMs) can generate them from natural-language requests by returning coordinates to a frozen segmenter. To that end, we convert masks from three mitochondria datasets into training examples, pairing images and instructions with centroid coordinates, then train LoRA adapters while freezing the MLLM backbone and microSAM. We find that Qwen3-VL reaches segmentation AP$_{50}$ $0.736$ after supervised fine-tuning and reward optimization, up from $0.247$ without adaptation, while automatic prompt generation (APG) achieves $0.773$. In addition, two other MLLMs improve, reaching or exceeding APG. When compared with a supervised centroid-heatmap detector that reaches AP$_{50}$ $0.904$ for this mitochondria task, Qwen3-VL more closely matches the annotated point set and instance counts. Moreover, training on two public datasets transfers to an unseen third, while training on all three transfers to an independent EM volume. Robustness tests show stable performance under unseen formulations of the natural-language request, while the coordinates can be reused by a second segmenter. To our knowledge, this is the first feasibility study of open-weight MLLMs as EM point generators, providing an inspectable, language-directed link between localization and mask decoding. --- ### LPA-CWM: A Learned Physical Adjudicator for Motion Reasoning with Counterfactual World Models Authors: Kunwei Wu, Xiang Liu, Guocai Yao, Junming Chen, Zhikang Chen, Min Zhang, Pengwei Wang, Sen Cui Published: 2026-09-12 Categories: cs.CV, cs.AI, cs.RO arXiv: https://arxiv.org/abs/2609.14073v1 Counterfactual world models (CWM) extract motion from pretrained video predictors by comparing factual and intervened predictions. However, responses generated under different target-frame masks vary in reliability, while uniform aggregation weights them equally. We formulate response aggregation as candidate reliability learning and propose LPA-CWM with a lightweight Learned Physical Adjudicator (LPA). Trained on dense MOVi-F trajectories, the 3.0M-parameter LPA compares visual context and response structure across an unordered candidate set to predict relative weights, while the CWM predictor and intervention generator remain frozen. The weighted responses undergo windowed localization and one paired re-evaluation to recover motion. We also introduce Completeness-aware Motion Correspondence (CMC), a ground-truth-anchored evaluation protocol that jointly measures localization, trajectory completeness, visibility, and continuity, counting missing predictions as failures on visible dynamic points. On the evaluated DAVIS and Kinetics subsets, LPA-CWM improves $\mathrm{DCA}_{\mathrm{avg}}$ over Uniform CWM by 60.0\% and 29.0\%, respectively, and also improves tracking accuracy under TAP-Vid First. A quick overview is available at https://LPA-CWM.github.io. ---