# Research Papers — 2026-09 · bundle page 156 of 266 > Full-text markdown of papers #7751–7800 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/156 [← Previous bundle](https://aipod.dev/papers/archive/2026-09/page/155.md) [Next bundle →](https://aipod.dev/papers/archive/2026-09/page/157.md) ### Derivative-Free Structured Updates for Muon Authors: Pengcheng Xie Published: 2026-09-15 Categories: math.OC, cs.LG arXiv: https://arxiv.org/abs/2609.17759v1 Muon updates matrix-valued neural-network parameters by orthogonalizing a gradient-based momentum matrix. Its reliance on derivatives limits its use when gradients are unavailable or unreliable. We develop a derivative-free framework that constructs Muon-style updates from structured finite differences. Four variants are considered: full entrywise recovery, random low-rank surrogates, basis-aligned rank-one probing, and direct structured search. Exhaustive basis-aligned probing is equivalent, up to positive scaling before ideal polar orthogonalization, to coordinate finite differences. Matrix-regression experiments show that random rank-one probing can reduce the number of function evaluations substantially, at the cost of less accurate updates. Controlled noisy-gradient experiments on regression and a neural network illustrate when accurate function values can compensate for an unreliable gradient oracle. A small CartPole study further examines orthogonal rank-one probes under a fixed episode budget. These results support structured probing as a practical option for selected black-box problems; they do not establish a general convergence guarantee or an advantage over accurate, inexpensive gradients. --- ### CALOS: Control-Affine Lyapunov On-manifold Safety Layer for Safe Deep Reinforcement Learning for Quadrotors Authors: Fabrizio Cesareo, Sebastiano Mengozzi, Nicola Mimmo, Andrea Acquaviva Published: 2026-09-15 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2609.17758v1 Deep Reinforcement Learning has demonstrated remarkable capability in quadrotor control, yet learned policies offer no guarantee of respecting safety constraints during training or deployment. We present CALOS (Control-Affine Lyapunov On-manifold Safety), a runtime safety layer that enforces attitude constraints on a quadrotor without modifying the underlying learning algorithm. CALOS formulates four tilt-angle inequalities and a Lyapunov descent condition as a single quadratic program whose solution is the minimum-norm correction to the nominal torque output of the policy. The quadratic program is solved exactly via active-set enumeration over the three-dimensional torque space, with a computational cost low enough to enforce constraints in real time across thousands of parallel simulation environments, as required by modern massively parallel Deep Reinforcement Learning training. Evaluated on trajectory-tracking tasks in NVIDIA Isaac Lab, CALOS reduces lateral tracking error by 55-60% relative to an unconstrained Proximal Policy Optimization baseline while achieving zero attitude-constraint violations on the training trajectory. By restricting exploration to safe regions of the state space, the safety layer also accelerates training convergence and improves data efficiency without producing suboptimal policies. --- ### Imitation Learning for Autonomous Driving in CARLA Authors: Jordy Kieto Published: 2026-09-15 Categories: cs.AI, cs.RO arXiv: https://arxiv.org/abs/2609.17757v1 Behavioral cloning trains a policy offline on expert demonstrations, but deployment is closed loop: each action affects the observations the policy receives next. We study how much closed-loop driving competence a compact multimodal policy can acquire from offline demonstrations in the CARLA simulator. The policy uses five-frame histories of RGB images, LiDAR, vehicle telemetry, and lane waypoints to predict throttle, brake, and steering at 20 Hz. Demonstrations were collected in three stages, ending with a systematic route-generation procedure that enumerates spawn points and feasible maneuvers and verifies completed autopilot routes. The released 1.36 million parameter policy was trained on 236,882 windows, representing about 3.3 hours of driving from 448 captures. The resulting policy drives autonomously for hours on training and held-out routes. In our runs, it did so without collisions and also transferred qualitatively to an unseen CARLA town with different road geometry. We also observed recovery from large trajectory deviations, although we do not claim systematic recovery without controlled evaluation. We report offline metrics and distinguish measured results from qualitative closed-loop observations. We release the code, trained checkpoint, ONNX model, data sample, and an evidence audit for the reported claims. --- ### Evolution of US Oral Political Language Authors: Jacques Savoy Published: 2026-09-15 Categories: cs.CL, cs.AI, cs.DL arXiv: https://arxiv.org/abs/2609.17755v1 The analysis of US political language is usually based on the written form (e.g. presidential addresses) or posts broadcasted on various social networks. Oral production, however, which is even more frequent, can better reveal the style and mode of thinking of the speaker. This study covers this mode of linguistic communication by considering 19 candidates from the presidential elections between 1960 to 2024. Our main research objectives are to disclose the main trends hidden in those presidential debates. Do we observe a clear simplification of the US political language over time? Does Trump have poor language compared to the other candidates? Do unusual stylistic features occur only with a single, specific president? Moreover, can we detect a pattern explaining the success or failure of some nominees? Over time, this study demonstrates a significant reduction in political language complexity, a decrease of the mean sentence length, and a noteworthy decline of complex terms. Moreover, the emotional tone increases over the decades, while logical and rational thinking tends to lessen. --- ### Beyond Performance Metrics: Uncertainty Mapping of Label Ambiguity in Fazekas Score Prediction Authors: Susanne Schmid, Johanna Ospel, Richard Frayne, Roberto Souza Published: 2026-09-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.17753v1 Reference labels used to train medical image classification models are not always as certain as they may appear, and this uncertainty has implications on performance metrics. In this study, we propose a framework to analyze model performance for periventricular Fazekas score prediction that goes beyond conventional metrics. The Fazekas score is an ordinal visual rating scale used to assess the severity of white matter hyperintensities and is known to be affected by inter-rater variability. While the best Fazekas score prediction model achieved a Matthews correlation coefficient (MCC) of 0.70, performance varied across data splits and loss functions, making interpretation of model capabilities difficult. Rather than interpreting epistemic uncertainty of a model's prediction as an isolated scalar value, our approach of uncertainty mapping relates uncertainty to its position within the learned feature representation. This highlights regions of class-boundary transitions where cases appear more ambiguous and misclassifications are more likely. It also identifies potential label disagreement, including low-uncertainty misclassified cases that expert review found to be inconsistent with the original reference Fazekas score. Therefore, uncertainty mapping allows model behaviour to be examined in relation to class separation and potential model-label disagreement. Loss function choice also influenced the uncertainty profile, with some models showing clearer class separation and more localized uncertainty in ambiguous regions than others. These findings suggest that uncertainty mapping for Fazekas score predictions can support model interpretation and targeted dataset review when reference labels are affected by ambiguity/ inter-rater variability. --- ### How to make effective use of domain experts for image classification? Authors: Dieu-Donné Fangnon, Diane Lingrand, Aurélie Liard, Marco Corneli, Antoine Pasqualini, Frédéric Precioso Published: 2026-09-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.17749v1 A lot of expectations have been put for years on integrating domain expert knowledge in image classification models. Several approaches have been explored, Concept Bottleneck Models (CBMs) opened up a new avenue of research leading to many variants, and more recently to Concept-based Embedding Models (CEMs). CBM consider binary encoding of each concept, while CEM expands this idea by embedding each concept through two vectors. However in real-life scenarii, domain experts' knowledge is usually organized in concepts determined by various attributes, each attribute encoded either with numerical values, or range of values, or binary values, or categorical values. In this work, we first finetune an image feature extractor for classifying attributes representing the downstream object classes, where the class attributes have been specified by experts under various encoding formats. A classification head is then learnt from these various attributes to categorize target objects. We experimentally show that it improves the classification for three datasets: Kaggle fish dataset, AWA2 and a more challenging new wood charcoal dataset. We then propose an automatic selection of potential missclassified data. In this second step, experts are asked for those data to eventually modify the predicted attributes in order to improve the classification. --- ### SAM-on-the-Curve: Sharpness-Aware Mode Connectivity for Robust Weight-Space Interpolation Authors: Alejandro Calatrava, Xu Zhang, Ren Wang Published: 2026-09-15 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.17748v1 Deep neural networks that are independently trained to similar performance can be connected by low-loss parametric curves in weight space, a phenomenon known as Mode Connectivity (MC). This geometric property underpins practical techniques such as weight averaging, model ensembling, and model merging. We argue that low-loss connectivity is an incomplete geometric criterion: it controls loss only along a one-dimensional trajectory while leaving the surrounding weight-space neighborhood unconstrained, so the optimized curve may traverse sharp ridges that become fragile under distribution shift. We therefore reformulate mode connectivity as a neighborhood-robust path optimization problem, seeking a curve whose entire local neighborhood maintains low loss. We propose Sharp Mode Connectivity (SMC), which applies a first-order sharpness-aware approximation to the resulting minimax functional, enforcing flatness along the entire curve rather than only on it. We derive a practical optimization algorithm for connectivity paths under this sharpness-aware objective. Under severe blur corruptions from CIFAR-10-C, SMC achieves up to 6.09\% absolute accuracy improvement over standard MC. Remarkably, SMC produces negative loss barriers, meaning that models obtained at interior points of the optimized path can outperform the average endpoint loss. These results, validated across ResNet-18, VGG16-BN, and ViT-Tiny on CIFAR-10 and ImageNet-100, establish path-wise flatness as a practical principle for robust weight-space interpolation. --- ### Is Trump's Vocabulary Poor? Vocabulary Richness Across Texts of Different Lenghts Authors: Dominique Labbe, Cyril Labbe, Jacques Savoy Published: 2026-09-15 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.17747v1 This study explores the vocabulary richness of oral political communication. A model explaining the lexicon growth is proposed by subdividing the whole vocabulary into terms generated by general and specialized glossaries. --- ### REVERSAL-BENCH: A Reversibility Axis and Reset Oracle for Measuring the Reset-Free RL Cliff Authors: Riyaaz Shaik, Chandru Venkataraman Published: 2026-09-15 Categories: cs.LG, cs.AI, cs.RO arXiv: https://arxiv.org/abs/2609.17745v1 A central goal of autonomous reinforcement learning is continuous policy training without external resets. However, existing paradigms largely depend on underlying environmental reversibility, a property absent in real world manipulation, where events such as pushing objects off tables or spilling granular substances cannot be undone. We introduce REVERSAL-BENCH, a benchmark that controls reversibility via a continuous parameter $ρ\in (0, 1)$ and provides a reset oracle, a ground-truth verification mechanism to test state recoverability across eight manipulation settings in five physics engines. Evaluating a broad spectrum of policy architectures, including standard actor-critic algorithms, safe RL, and specialized reset-free frameworks, reveals a sharp reversibility cliff: reset-free agents are consistently absorbed into irrecoverable states as $ρ$ increases, whereas episodic agents maintain steady learning. We see this failure mode across autonomous reset-free baselines and constrained RL. Because reset-free agents lack external resets, any transition into an irrecoverable state results in permanent absorption, leaving the agent trapped where further learning halts. We show that this absorption phenomenon persists in full physics simulations under learned manipulation policies. By evaluating against geometrically identical reversible counterparts, we confirm that this breakdown is causally driven by irreversibility rather than obstacle complexity. We release the benchmark suite, a large multi-simulator dataset labeled with recoverability and a reset oracle. We also evaluate a safety shield that intervenes before irreversible failures occur, showing that while recoverability can be predicted accurately, active recovery primarily succeeds only when the agent can physically steer clear of the trap --- ### Geometry-Driven Shadow Harmonisation for Composited Faces: A Multiplicative, Albedo-Preserving Relighting Pipeline Authors: Vijesh KP Published: 2026-09-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.17740v1 Face swapping and face compositing pipelines routinely produce a face that is geometrically well aligned but photometrically implausible: the donor face carries flat, near-frontal studio illumination while the host body and background carry directional scene light. Most existing remedies re-synthesise the face through colour transfer, neural relighting, or inverse rendering, and therefore risk altering identity, skin tone, and texture. We present a conservative alternative: geometry-driven form-shadow injection. The pipeline never repaints the face. It estimates a per-pixel gain field $g\in(g_{\min},1)$ from a rasterised 3D face proxy and multiplies it channel-uniformly onto linear RGB, so the operator can only darken and cannot shift chromaticity. A dense landmark mesh is rasterised into a depth buffer, from which we derive surface normals, a cavity term, and screen-space cast shadows. Key-light direction is estimated from host-side cues (body, background, hair halo); on-face cues are downweighted because they recover the donor's lighting. Shadow magnitude is not matched to the host: it is set by a three-parameter transfer $(τ,σ,g_{\min})$. The shading field is divided by its 75th percentile over skin, then gated, scaled, clamped, smoothed, and re-clipped inside a feathered, skin-gated face mask. On an analytic face heightfield, the default $(τ,σ,g_{\min})=(0.90,0.45,0.82)$ modifies 56.5% of face pixels with mean gain 0.938 (0.890 on modified pixels) and drives 3.4% of pixels to the floor. Hue invariance is a corollary of the operator. We analyse the transfer in closed form, ablate its parameters, and discuss failure modes of a monotone, darkening-only formulation, including double-shadowing of non-flat donors. --- ### Similarity Pairing with Energy Mover's Distance for Self-Supervised Pre-Training at the LHC Authors: Ho Fung Tsoi, Dylan Rankin Published: 2026-09-15 Categories: hep-ex, cs.LG arXiv: https://arxiv.org/abs/2609.17738v1 Many self-supervised methods for training foundation models at the Large Hadron Collider (LHC) rely on data augmentations to encourage the model to embed events into a representation space invariant to certain physical or detector symmetries. A common challenge arises from the large freedom in choosing a proper set of augmentations on which downstream performance depends. The implementation of augmentations involves either modifying existing events, potentially breaking the event fidelity, or simulating more event variants, which is computationally intensive. In this work, we present a data-driven method of pairing events by their similarity via the energy mover's distance (EMD), which measures how similar two events are in terms of the work required to transform one into the other. With this approach, distinct events are sampled and matched by their similarity to serve as views for learning invariance, keeping the physics content of each event intact without handcrafted distortions. We demonstrate this augmentation-free pairing method by pre-training on QCD jets via self-distillation and show that it can yield semantic jet embeddings with downstream discrimination power comparable to or better than an augmentation-based baseline. --- ### Machine learning kinetics from molecular dynamics data Authors: Jonathan Weare, Aaron R. Dinner Published: 2026-09-15 Categories: physics.comp-ph, cond-mat.stat-mech, cs.LG, physics.chem-ph, physics.data-an arXiv: https://arxiv.org/abs/2609.17736v1 Most molecular transitions occur on timescales far beyond direct molecular dynamics simulations. The committor, the probability that a configuration reaches a product state before a reactant state, is a central kinetic statistic, providing a mechanism-independent reaction coordinate and a foundation for transition path theory and the calculation of rates. This review surveys modern approaches for estimating the committor and related kinetic statistics from molecular simulations, with an emphasis on self-supervised methods that learn solutions of their defining dynamical equations rather than relying on labeled shooting data. We develop a common operator viewpoint connecting generator-based partial differential equations, variational principles, Markov state models, dynamical Galerkin approximation, and neural networks. Empirical and theoretical evidence points to the efficiency of these methods. We provide theoretical and practical guidance for realizing their full potential in applications, including strategies for treating non-Markovian effects and for sampling. We conclude by identifying opportunities for further research, including connections to reinforcement learning and generative modeling. --- ### A Systematic Evaluation of the COTQ Provincial Land Cover Product: Structural Consistency, Spectral Separability, and Relative Positioning Against ESA, ESRI, and Google Products Authors: Étienne Clabaut, Samuel Foucher, Yacine Bouroubi Published: 2026-09-15 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.17731v1 High-resolution land use and land cover (LULC) products derived from Sentinel-2 imagery are widely used for environmental monitoring and land management, yet their performance can vary across regions with complex ecological gradients and heterogeneous surface conditions. In Quebec, these limitations motivated the development of a provincial 10-m land-cover product, the COTQ, designed to support annual monitoring of land occupation and soil artificialisation. This study presents a systematic evaluation of the COTQ product relative to three global 10-m LULC datasets: ESA WorldCover, ESRI LandCover, and Google DynamicWorld. This paper does not introduce a new mapping methodology but focuses on analysing the behaviour and consistency of the COTQ using complementary evaluation approaches. All products are harmonized under a common legend and compared using structural indicators (object-size distributions, shape complexity, Adjusted Rand Index, and Intersection over Union), spectral separability metrics derived from Sentinel-2 reflectance data, and a targeted photo-interpretation of disagreement areas. The analysis is conducted over eight Sentinel-2 tiles selected to represent the main bioclimatic domains of Québec, from temperate and boreal forests to northern tundra environments. The results show that the COTQ exhibits structural and spectral characteristics most similar to ESA WorldCover among the reference global products, while revealing systematic differences linked to class definitions and thematic priorities, particularly for urban areas, wetlands, and rocky or cryptogamic surfaces. This multi-criteria evaluation provides an objective characterization of the COTQ product and clarifies its relative positioning with respect to existing global land-cover datasets for operational land monitoring in Québec. --- ### FAME: An FPGA-Based Platform for Approximate Multipliers Evaluation with Pattern-Guided DNN Retraining Authors: Rappy Saha, Nima Amirafshar, Jude Haris, Nima Taherinejad, José Cano Published: 2026-09-15 Categories: cs.AR, cs.LG, cs.PF arXiv: https://arxiv.org/abs/2609.17730v1 Approximate multipliers can reduce hardware area and energy consumption in Deep Neural Network (DNN) inference; however, they introduce computational errors. Assessing the accuracy of numerous approximate multiplier designs across diverse DNN models and large-scale datasets remains challenging due to prohibitive evaluation times. This overhead primarily stems from the slow emulation of approximate multiplier behavior using look-up tables (LUTs) on CPU and GPU platforms. Moreover, the resulting accuracy degradation must be carefully quantified and, if necessary, mitigated (e.g., through retraining), further increasing the overall evaluation cost. To address these challenges, we propose FAME, an FPGA-based platform for evaluating approximate multipliers. The platform exploits the reconfigurable logic of Field-Programmable Gate Arrays (FPGAs) to implement approximate multipliers directly in hardware, eliminating the need for LUT-based emulation on CPU/GPU platforms and thereby enabling efficient DNN inference while significantly reducing evaluation time on large datasets. Furthermore, we introduce a pattern-guided DNN retraining technique to mitigate accuracy degradation induced by approximate multipliers. Specifically, retraining is guided by multiplier-specific patterns to effectively recover potential accuracy losses. We evaluate FAME using two DNN models, ResNet-18 and MobileNetV2, on the ImageNet dataset across 27 approximate multipliers. During inference, our approach achieves up to a 3.47x speedup in approximate multiplier evaluation compared to prior LUT-based emulation methods. Furthermore, the proposed retraining technique improves accuracy by up to 65.5% over existing retraining approaches for the evaluated multipliers. The code is publicly available at: https://github.com/gicLAB/FAME --- ### Self-Supervised Learning for Robust Resonance Mass Regression in Cascade Decays Authors: Ho Fung Tsoi, Alex Yang, Luis Felipe Gutierrez Zagazeta, Shion Chen, Dylan Rankin Published: 2026-09-15 Categories: hep-ex, cs.LG arXiv: https://arxiv.org/abs/2609.17726v1 Reconstructing the mass of a heavy resonance from its decay products with missing energy is one of the central tasks that directly determine the sensitivity in new physics searches at collider experiments. Supervised learning approaches to this problem often struggle to generalize well due to the presence of various systematic uncertainties and distribution shifts. Exhausting all possible variations in the labeled data can be very compute-intensive, while a failure of the model to generalize can corrupt the reconstructed resonance widths that are critical in peak-hunting analyses. In this work, following the foundation model paradigm, we use a self-supervised approach to pre-train a transformer encoder with VICReg to learn an embedding invariant to various corruptions, then fine-tune it for mass regression on a heavy resonance with masses ranging from 2.5 to 6.5 TeV and a SUSY-like cascade decay into an eleven-body final state. We show that the pre-trained model reconstructs sharper resonance peaks and has a more stable performance under various realistic corruptions, compared to a supervised model of the same architecture trained on the same augmented data from scratch. --- ### Decoding Extrahepatic Targeting of Lipid Nanoparticles with Interpretable Machine Learning Authors: Asal Mehradfar, Mohammad Shahab Sepehri, Owen Antholine, Varun Shankar, Glen S. Kwon, Salman Avestimehr, Morteza Rasoulianboroujeni Published: 2026-09-15 Categories: q-bio.QM, cs.CE, cs.LG arXiv: https://arxiv.org/abs/2609.17721v1 Lipid nanoparticles (LNPs) have transformed RNA medicine, yet their clinical utility remains constrained by predominant hepatic accumulation after systemic administration. Redirecting LNPs to extrahepatic tissues requires understanding of how lipid chemistry and formulation composition jointly govern in vivo biodistribution. Here, we develop an interpretable machine learning framework to predict hepatic versus extrahepatic LNP accumulation and identify molecular design rules for extrahepatic RNA delivery. A literature-derived dataset of 476 intravenous LNP formulations was curated from 81 studies, integrating formulation composition, lipid chemical structures, and IVIS-based biodistribution profiles. Standardized SMILES representations of ionizable lipids, helper lipids, sterols, PEGylated or polymer-conjugated lipids, additional lipids, and polymer repeat units were converted into RDKit Expert descriptors and combined with formulation-level variables to generate an 808-dimensional feature representation. Logistic regression, random forest, and XGBoost achieved ROC-AUC values of 0.839, 0.866, and 0.874, respectively. SHAP-based interpretation and consensus feature ranking revealed that ionizable-lipid descriptors dominate biodistribution prediction, while formulation composition, particularly ionizable lipid, sterol, and PEGylated/polymer-conjugated lipid fractions, contributes substantially. The top 20 consensus features retained nearly all predictive information in tree-based models. The most informative features implicated electrotopological surface properties, charge- and hydrophobicity-weighted surface areas, molecular topology, and amide/alkyl structural motifs as drivers of extrahepatic accumulation. This study establishes an interpretable, data-driven strategy for decoding LNP biodistribution and provides actionable design principles for engineering LNPs beyond the liver. --- ### One Size Does Not Fit All! Dynamic Retriever and Generator Selection for RAG Authors: Neeraj Anand, Payel Santra, Partha Basuchowdhuri, Debasis Ganguly, Sumit Bhatia Published: 2026-09-15 Categories: cs.IR, cs.AI arXiv: https://arxiv.org/abs/2609.17709v1 Retrieval-Augmented Generation (RAG) systems typically employ fixed retriever and generator configurations across queries, despite substantial differences in query complexity and information needs, leading to inefficient allocation of computational resources. While retrieval and generation adaptivity have been studied independently, their joint effect on end-to-end RAG performance remains underexplored. We systematically analyze how retriever and generator complexity interacts across factoid and multi-hop question answering (QA), including bridge and composition reasoning tasks. Our analysis shows that stronger retrieval generally yields larger gains than increased generation effort, but both exhibit diminishing and non-monotonic returns, indicating that higher-complexity configurations are not uniformly better across queries. Motivated by these findings, we introduce DRAG, a query-adaptive framework for selecting retriever-generator configurations. We first propose DRAG$_\text{QPP}$, a training-free routing approach that uses Query Performance Prediction (QPP) signals to guide retriever selection and perplexity-based measures over retrieved context to guide generator selection. We further introduce DRAG$_\text{SFT}$, a supervised routing approach that fine-tunes an LLM to jointly predict retriever-generator configurations. Across three LLM families and four QA benchmarks, \qpprag~achieves performance comparable to strong static RAG baselines while substantially reducing inference latency, whereas DRAG$_\text{SFT}$ consistently improves effectiveness over static and training-free adaptive baselines. Overall, DRAG demonstrates that jointly adapting retrieval and generation achieves a more favorable effectiveness-efficiency trade-off than static RAG pipelines. --- ### Confidence Comes from Experience: Experiential Confidence Estimation from Reasoning to Agents Authors: Caiqi Zhang, Xiaochen Zhu, Chengzu Li, Yulong Chen, Dharshan Kumaran, Nigel Collier Published: 2026-09-15 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2609.17708v1 Reliable confidence estimation is increasingly central to the trustworthy deployment of language models: a calibrated estimate of the probability that an output is correct decides what to ship, what to escalate, and what to retry. Existing confidence estimators, however, share one design premise: they only read the current inference process, either by introspecting on it, scoring its token probabilities, or resampling it. We argue that the current inference is not a sufficient basis for confidence. We propose XConf (eXperiential Confidence): estimating confidence together with the model's accumulated experience. The experience is stored as a record of the model's own graded past episodes, each holding the task, the model's reflection, its stated confidence, the outcome, and a lesson written once the grade arrived. Given a new task, XConf's Recall stage retrieves past episodes on similar tasks met with a similar stated confidence, and reads off their historical success rate; its Reflect stage shows the model this record, has it name its recurring failure mode, and restate a confidence now informed by its own track records. Our estimator is format-general, requiring no logit access or weight updates, and costs only one answer generation. Across nine benchmarks spanning reasoning, coding, multimodal QA, and interactive agents, and four models from three families, XConf beats or matches ten-sample self-consistency in discrimination (AUROC) on 23 of 24 comparisons, with much lower calibration error (ECE), at a tenth of the generation cost. Used for selective prediction, abstaining on the 10% least-confident episodes raises the delivered success rate by up to 8.7 points on agent tasks. We therefore see experiential confidence estimation as a new paradigm for future general-purpose confidence estimation. --- ### NeMo Data Designer: An Extensible Framework for Multimodal Synthetic Data Generation Authors: Johnny Greco, Nabin Mulepati, Andre Manoel, Eric Tramel, Kirit Thadaka, Mike Knepper, Dhruv Nathawani, Dane Corneil, Yev Meyer, Alex Watson, Maarten Van Segbroeck Published: 2026-09-15 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.17699v1 We present NeMo Data Designer (NDD), an open-source, general-purpose framework for multi-modal synthetic data generation (SDG). Designed to be intuitive to use, NDD provides a declarative configuration format in which human and/or agent users define each dataset column, with column types spanning text, code, structured outputs, images, embeddings, and statistical samplers that are explicitly configured to steer dataset diversity. Additional column types and functionality can be introduced using the framework's flexible plugin system. NDD's configuration is an inspectable artifact, supporting workflow sharing and reproducibility. SDG is an inherently iterative process. NDD therefore builds a preview-and-revision loop into its core workflow, allowing users to generate and inspect a small number of records, refine the specification, and rerun generation at full scale. At runtime, NDD resolves dependencies, schedules calls to user-provided model endpoints, and retries failed requests. We describe NDD's architecture and programming model and present case studies spanning structured, agentic, multimodal, and domain-specialized tasks, including datasets used in Nemotron model development and in production enterprise deployments. --- ### Composite-Gradient Learning for Shared Control Authority Between Deep Reinforcement Learning and Model Predictive Control Authors: Giray Önür, Azita Dabiri, Bart De Schutter Published: 2026-09-15 Categories: cs.LG, eess.SY arXiv: https://arxiv.org/abs/2609.17697v1 Integrated deep reinforcement learning (DRL) and model predictive control (MPC) methods are increasingly used to control autonomous systems by combining their complementary capabilities. DRL learns control policies through interaction with the environment. MPC uses a system model to optimize control inputs while accounting for constraints. In DRL-MPC frameworks with shared control authority, both the DRL agent and the MPC controller each determine part of the control inputs. However, common learning formulations treat MPC as part of the environment and therefore do not explicitly account for MPC's contribution to control or its interaction with the DRL agent. This paper proposes a novel composite-gradient learning (CGL) method that integrates the MPC controller into the learning process by representing the DRL and MPC control inputs as a joint action and accounting for their interaction when updating the DRL agent during training. CGL is evaluated on two multi-class freeway traffic networks with different strengths of interaction between the DRL and MPC control inputs and it is compared with alternative methods that treat MPC as part of the environment or that only partially incorporate MPC into learning. The results show that CGL offers limited benefit under weak interaction, but learns higher-performing control policies than the alternative methods in a subset of training runs under strong interaction, although the average control-performance gains remain modest. --- ### GVD: Governed Versioning and Deduplication for Document Repositories Authors: Mohammadreza Sediqin, Shivali Dalmia, Sumukha Thoppanahalli, Abhishek Mukherji Published: 2026-09-15 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.17696v1 Document repositories evolve continuously. Guidelines and policies are revised, superseded, and re-uploaded, so the same content recurs in different wording and newer versions refine or contradict earlier ones. These inconsistencies belong to the growing collection rather than to any single document, yet existing work treats versioning, duplicate detection, and contradiction detection as isolated pairwise tasks and stops once a pair is labeled. We present GVD (Governed Versioning and Deduplication), a framework that unifies cross-document version linking with rule-level conflict resolution under an auditable update policy. Incoming documents are assigned to version families through bidirectional rule alignment, and their rules are compared against the family memory to identify duplicates, contradictions, asymmetric refinements, and new knowledge, with Counterfactual Span Probing (CSP) resolving related pairs that inference misclassifies as neutral. Relation-specific policies suppress duplicates and escalate only consequential changes for review, retaining version lineage as an audit trail. The pipeline runs fully locally, with no large language model. On 120 enterprise documents processed as 140 ingestions across 59 version families, GVD reaches an F1 of 0.97 for version-family construction and 0.94 for rule-level consistency, with CSP raising rule consistency from 0.90 to 0.94. --- ### GraphEcho: Structural Redundancy and Evidence Provenance in LLM Graph Agents Authors: Sikun Wang, Yixi Zhou, Lei Fan, Fan Zhang Published: 2026-09-15 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.17695v1 A large language model (LLM) agent can follow more graph paths without acquiring more independent evidence. GraphEcho tests whether agents mistake these repeated encounters for additional corroboration. The benchmark varies path counts and evidential origins while holding evidence content fixed, and evaluates both judgments and active exploration. Controlled synthetic experiments reveal model-dependent judgment shifts, but redundant supporting paths increase the share of repeated walks across all evaluated frozen agents. Provenance-aware post-training (PAPT) reduces revisits and improves synthetic accuracy, yet covers fewer distinct sources. On scientific claims, it continues to reduce repetition while accuracy declines. These findings expose a gap between efficient exploration and effective evidence use: an agent can learn to stop repeating itself while overlooking information it needs. GraphEcho provides a controlled way to evaluate both what graph agents conclude and whether their exploration reaches distinct evidential sources. --- ### Accelerating Diffusion Sampling via Speculative Draft Trees Authors: Marcello Bullo, Yanxiao Liu, Öykü Sıla Güner, Arpan Mukherjee, Deniz Gündüz Published: 2026-09-15 Categories: cs.LG, cs.AI, stat.AP arXiv: https://arxiv.org/abs/2609.17691v1 Speculative sampling accelerates diffusion model generation by drafting inexpensive candidate states and correcting them under a coupling that preserves the target distribution exactly, reducing the number of expensive target evaluations. Existing diffusion samplers, notably those based on reflection maximal coupling, are topologically constrained: their lookahead drafts form a chain graph, a single linear sequence, which inherently limits the acceptance rate per target evaluation. We connect speculative sampling in diffusion models to relative entropy coding (REC). This perspective shows the lookahead need not be linear and motivates our central contribution, draft trees, which enrich the candidates considered per round and lower the target function evaluations. We further adopt greedy rejection sampling, an REC algorithm, as the draft-target coupling, improving acceptance while guaranteeing exact target samples. Experiments across diverse target and draft models demonstrate up to 8.3% acceleration over the reflection coupling baseline in practical settings. --- ### CapMem: A Benchmark for Caption-Based Episodic Memory in Egocentric Video Authors: Dingli Liang, Yiqiao Xie, Yukai Huang, Zhaokai Wang, Weitong Cai, Guangwen Feng, Jifei Song, Zhensong Zhang, Hang Zhang Published: 2026-09-15 Categories: cs.AI, cs.CV arXiv: https://arxiv.org/abs/2609.17688v1 Wearable assistants require episodic memory over egocentric video, yet current vision-language models face bounded frame budgets, growing visual-token costs, and long-context retrieval failures. Under these practical constraints, we study whether textual captions can serve as reusable episodic memory. We define the Episodic Memory Video Caption QA task and introduce CapMem, a human-annotated benchmark with 75 videos totaling 33.7 hours, and 1,000 multiple-choice questions across 16 scenarios. On long videos (>20 min), full-coverage CaptionQA with 30s and 60s caption windows outperforms direct VideoQA for 10/12 and 8/12 models, respectively. On the same video subset, a matched-frame control across six Qwen models retains mean accuracy gains of 3.22 and 2.55 points, respectively. Our caption-guided retrieve-and-verify harness further improves accuracy by up to 5.3 points. These results support the effectiveness of caption memory for episodic reasoning over long egocentric video. --- ### The Missing "I Don't Know": Why Three Reasoning-Reliability Findings Converge on Calibrated Abstention Authors: Srijith Ravikumar Published: 2026-09-15 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.17686v1 Three recent results describe what look like unrelated LLM reliability problems. Yin et al. (2026) show reasoning RL collapses tool-reliability representations. Suleymanov et al. (2026) show that under safety-constrained generation, large models rewrite flagged spans while small models truncate. Bastounis et al. (2024) prove any consistent-reasoning system without an implicit "I don't know" function must hallucinate infinitely often on broad problem classes. We argue these findings converge on a single intervention: calibrated abstention is what each independently identifies as the missing capability, even though the unavailability they document, a capability gap, a policy gap, and a recursion-theoretic gap, has a different source in each case. Honesty post-training has narrowed the gap in deployed models, but principled closure of the class Bastounis identifies requires a calibrated abstention function whose training signal at the leaderboard level is absent: dominant benchmarks assign zero reward to decline, so the leaderboard gradient that would select for the function does not exist. We propose four changes to evaluation: triple-scoring, abstention-rate reporting, capability-stratified evaluation, and mandatory calibration metrics. Benchmark reform is necessary, not sufficient, for closing the gap the theorem identifies. --- ### DSD: Learning Diverse and Reusable Motor Skills via Diffusion Skill Discovery Authors: Sun Woo Kim, Xue Bin Peng Published: 2026-09-15 Categories: cs.LG, cs.GR arXiv: https://arxiv.org/abs/2609.17682v1 Humans efficiently learn new tasks by reusing a rich repertoire of motor skills across different goals and contexts. A similar strategy can also be used to enable simulated characters to efficiently perform new tasks by leveraging reusable motor skills. To support a wide range of downstream tasks, the learned repertoire should be diverse, consisting of distinct behaviors as well as spatial and temporal variation within each behavior. A commonly used method for learning diverse skills is by maximizing the mutual information between skill latents and the states produced by a policy. The marginal state entropy promotes broad behavioral coverage, while the conditional entropy encourages consistent behaviors from each latent. However, directly estimating the marginal state entropy is intractable in high-dimensional control problems. Prior methods therefore rely on indirect latent-space approximations or coarse estimators of the state distribution. These approximations may not effectively promote broad coverage of the state space, resulting in skills with limited behavioral diversity and reduced utility for downstream tasks. In this work, we propose Diffusion Skill Discovery (DSD), a skill discovery method that uses a diffusion model to approximate the entropy gradient of the policy-induced state distribution through score matching. The resulting objective encourages the discovery of skills that produce a broader range of behaviors for high-dimensional humanoid control. The learned skills are reused in two downstream control settings: hierarchical control with a task-specific high-level policy and zero-shot control through latent selection from offline trajectories. Our experiments show that DSD discovers a broader repertoire of reusable motor skills than prior skill discovery methods, leading to the emergence of complex and agile behaviors that can be reused across downstream tasks. --- ### Efficient Robust Learning at the Information-Theoretic Limit Authors: Adam R. Klivans, Konstantinos Stavropoulos, Sergei Tikhonov, Arsen Vasilyan Published: 2026-09-15 Categories: cs.DS, cs.LG, stat.ML arXiv: https://arxiv.org/abs/2609.17655v1 In an important recent work, Blanc (2026) gave an algorithm for robustly learning Boolean concept classes with respect to a fixed distribution that outputs a (randomized) classifier achieving the optimal error of $η+ \varepsilon$ where $η$ is the noise rate. In contrast, it is well known that deterministic hypotheses cannot achieve error less than $2η+ \varepsilon.$ Blanc's algorithm is computationally inefficient, and the main problem left open in his work is to find a polynomial-time algorithm given access to an oracle for empirical risk minimization (ERM). In this paper, we resolve this problem and give such an algorithm. Perhaps surprisingly, our techniques make crucial use of various types of no-regret learners. Additionally, we give an efficient algorithm (no ERM oracle required) for robustly learning any function class that admits sandwiching polynomials with respect to hypercontractive distributions. As one consequence, we give the first polynomial-time algorithm for robustly learning a halfspace with respect to Gaussian marginals that achieves error $η+ \varepsilon$ for any constant $\varepsilon$. --- ### Regularized Least Squares Training of Quadratic Neural Networks with Applications to System Identification Authors: Luis Rodrigues, Zachary Yetman Van Egmond, Mohammad R. Amiri Fard Published: 2026-09-15 Categories: cs.LG, eess.SY arXiv: https://arxiv.org/abs/2609.17654v1 This paper proposes a least squares approach for the training of quadratic neural networks with regularization. The proposed methodology yields a lower bound on the solution of the training optimization problem for the case where the regularization coefficient is positive. Moreover, it yields closed-form expressions for the approximate solution and its sensitivity The lower bound is tight and the approximate solution is the optimal solution when the regularization coefficient is zero. Having a closed-form expression for the weights reduces considerably the computational time when compared with iterative numerical methods such as backpropagation that can get stuck in local minima. The proposed approach has three main contributions, namely, (i) it yields an analytical expression for the weights, (ii) an analytical expression for the sensitivity of the weights to errors in the data is also provided, (iii) it establishes a connection between the optimization to compute a lower bound and nuclear norm minimization. The proposed least squares training is successfully applied to a nonlinear system identification example where the proposed lower bound is compared with the optimal value. --- ### Reflect, Revise, Reuse: Training-Free Skill Evolution for GUI Agents Authors: Bofan Chen, Boxuan Zhang, Fei Tang, Zhengxi Lu, Yong Du, Tongbo Chen, Weiming Lu, Jun Xiao, Yueting Zhuang, Yongliang Shen Published: 2026-09-15 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.17653v1 GUI agents execute long-horizon tasks on dynamic graphical user interfaces, where pop-ups, delayed loads, and relocated widgets routinely invalidate plans fixed before execution. Recent agent-skill frameworks encapsulate reusable procedural knowledge to mitigate this, yet existing skill designs are largely developed without targeting GUI execution dynamics and treat skills as static artifacts produced before deployment rather than living procedural knowledge that improves through it. We argue that what GUI agents need is not better static skills, but skills that can be revised from execution feedback at deployment time, without additional training. We propose \textbf{EvoSkill-GUI}, a training-free framework in which each skill is a structured multi-file package containing retrieval metadata, executable plans, backup localization, failure-recovery rules, accessibility utilities, and failure cases. EvoSkill-GUI operates through a \textbf{\emph{reflect-revise-reuse}} loop: the executor performs instant in-rollout revisions, an isolated critic diagnoses failed trajectories under strict information isolation, and the executor edits specific skill files through a restricted tool interface. Across MobileWorld, AndroidWorld, and OSWorld, three mainstream GUI benchmarks spanning mobile and desktop platforms, EvoSkill-GUI consistently improves multiple base models without any training, with maximum gains of $+16.2\%$, $+6.0\%$, and $+10.5\%$ respectively, and evolved skill libraries continue to benefit related tasks rather than being rebuilt from scratch. Our code is available at https://github.com/ZJU-REAL/EvoSkill-GUI. --- ### Fathom: Per-Query Read Depth for Sparse Decoding over Offloaded KV Caches Authors: Vivek Kalyanarangan Published: 2026-09-15 Categories: cs.LG, cs.CL, cs.DC arXiv: https://arxiv.org/abs/2609.17652v3 When agentic sessions run to a million tokens with many sessions resident at once, the KV cache and the index that ranks it live in host memory, and the scan that ranks all n keys for a top-k step becomes the traffic that bounds decoding. We present Fathom, a key scan in which each query decides how many bits of each key channel to read. The 4-bit K cache is stored channel-major as bit planes, so a prefix of t planes is exactly the channel's t-bit quantizer, and the query spends its bit budget by reverse water-filling over the variance-weighted importance of its channels. At one million tokens on Qwen3-8B a decode step is 1.67x faster in GPU time than with the 136-bit scans of Double Sparsity, Loki and SparQ r=32, and in the same GPU time as SparQ's 68-bit read (r=16) Fathom reads 18% fewer bytes with lower attention error on six of seven model and context settings. On RULER-style tasks every per-token scan matches exact top-k decoding, and on real coding-agent sessions Fathom reaches the step agreement of the most accurate 136-bit scan at 92 bits. The store is the 4-bit K copy a quantized serving stack already holds, and the method is not faster when the index is resident in GPU memory. --- ### Robust and Efficient AI Frameworks for Scalable Material Design and Property Prediction Authors: Kishalay Das Published: 2026-09-15 Categories: cond-mat.mtrl-sci, cs.LG arXiv: https://arxiv.org/abs/2609.17646v1 This thesis develops robust and efficient AI frameworks for accelerating crystalline materials discovery by addressing both major stages of the materials-design pipeline: crystal property prediction and crystal structure generation. Motivated by the high computational cost of Density Functional Theory (DFT) and the limited availability of labeled materials data, the thesis explores graph representation learning, pretraining, multimodal learning, and generative modeling for scalable materials design. For property prediction, the thesis first introduces CrysXPP, which learns transferable crystal representations through unsupervised graph autoencoding, reducing dependence on large property-labeled datasets. It then proposes CrysGNN, a large-scale self-supervised graph pretraining framework that captures atomic connectivity, chemical attributes, and global structural information and transfers this knowledge to downstream property predictors through knowledge distillation. CrysMMNet further enriches crystal representations by jointly modeling graph structure and textual descriptions, thereby incorporating both local chemical and global structural knowledge. For crystal generation, the thesis introduces TGDMat, a text-guided joint diffusion framework that jointly models lattice parameters, atomic types, and atomic coordinates while incorporating textual structural knowledge during denoising. This enables the generation of more valid and stable periodic materials while also supporting conditional generation from natural-language descriptions. Overall, the thesis establishes a unified AI-based framework for data-efficient property prediction and controllable crystal generation, demonstrating how graph learning, multimodal representations, and generative models can reduce computational cost and improve the scalability of materials --- ### Rethinking Domain Specialization for Open-Ended Scientific Reasoning in Astronomy Language Models Authors: Vanessa Lama, Sanjay Das, Emily Herron, Yuan-Sen Ting, Tijmen de Haan, Junqi Yin, Tirthankar Ghosal, Feiyi Wang Published: 2026-09-15 Categories: astro-ph.IM, cs.AI, cs.CY, cs.LG arXiv: https://arxiv.org/abs/2609.17644v1 Domain-specialized language models are widely used for scientific question answering, but stronger general-purpose systems raise a sharper question: when does domain-specific fine-tuning remain valuable for open-ended scientific reasoning? We study this in astronomy with a curated QA benchmark from publicly available 2017--2026 Olympiad-style materials. The free-response subset contains 300 questions, including 204 text-only and 96 image-linked examples. We compare open-weight and API-served general-purpose, multimodal, and astronomy-specialized models using judge-based correctness and complementary reference metrics. Strong general-purpose models establish the highest correctness baseline in this testbed, while analyses of metric agreement, judge sensitivity, benchmark composition, and modality reveal variation not captured by a single leaderboard. These results motivate treating domain specialization as a task- and deployment-dependent property and highlight the role of domain-specific evaluation in determining which models, capabilities, and evaluation criteria are appropriate for scientific workflows. --- ### Scaling Articulated Rationales for MLLM-based Recommendation Authors: Haoke Xiao, Yueyang Liu, Yuhui Zhang, Xiang Chen, Yufei Liu, Jia Xu, Yalong Guan, Xiaolan Zhu, Xiaoyu Zhang, Shijun Wang, Shuang Yang, Zijie Meng et al. Published: 2026-09-15 Categories: cs.IR, cs.AI arXiv: https://arxiv.org/abs/2609.17639v2 We presented SARA, an industrial framework that transforms sparse articulated user rationales into scalable recommendation signals. Its data engine curates questionnaire responses into SARA-HQ, providing explicit preference supervision for aligning SARA-7B through SFT and Quality-Refining DPO. This alignment extends rationale generation from $86{,}564$ questionnaire-covered authors to the full $10$M-author space. SARA-Ranker translates the generated positive and negative rationales into features for user--author interaction modeling and negative-feedback history modeling, connecting articulated reasons to production ranking. Evaluation on unseen authors demonstrates that SARA-7B generates more specific, relevant, and grounded rationales than the evaluated general-purpose MLLMs. On top of a strong industrial ranking baseline with multimodal features, separate online A/B tests show that positive-rationale integration increases watch time by $0.99\%$, while negative-rationale integration reduces Hate feedback by $8.16\%$. Daily refresh and more than $30$ days of production deployment further demonstrate the operational feasibility of the approach. These findings establish articulated rationales as a useful complement to behavioral and content signals, and demonstrate a practical role for MLLMs in scaling sparse human explanations into preference information that improves industrial recommendation. --- ### Lecture notes on Physics Informed Neural Networks, Neural Operators, and their applications Authors: Alessandro Bombini Published: 2026-09-15 Categories: cs.LG, cs.AI, cs.CE, physics.comp-ph arXiv: https://arxiv.org/abs/2609.17638v1 This is the set of lecture notes for the PhD course \href{https://www.unibz.it/en/faculties/engineering/phd-computer-science/study-course-offering/2025/36967}{\textit{Physics Informed Neural Network}, held at the University of Bozen/Bolzano} in the academic year 2025/2026. The goal of the course was to introduce the concept of Physics Informed Deep Neural Networks (PINN) and Neural Operators (NOs), discuss their implementation from scratch in PyTorch and using advanced ad-hoc developed open-source libraries such as NVIDia PhysicsNeMo to address real-world problems in various fields (engineering, physics, petroleum reservoir). We discuss recent topics such as Mixture-of-Models, Fourier Neural Operators, Physics-Informed Kolmogorov-Arnold Networks (PIKANs) and Fourier Neural Operators. --- ### What You Can't See Is Still What You Learn: A Preregistered Sixty-Society Confirmation That Evidence Masking Drives Compositional Generalization Authors: Narcis Marincat Published: 2026-09-15 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.17637v2 Restricting what a module can read may improve what a system learns to compute. We test this in a preregistered confirmation with sixty four-cell systems sharing a frozen language-model backbone and communicating through learned continuous packets. Five conditions vary evidence masking, ownership markers, and replacement of foreign evidence with neutral filler (task-irrelevant text of the same token length), across six initialization clusters, each with two data orders, on one fresh task world. With markers available in both regimes, masking improved accuracy on held-out two- and three-operation compositions by median paired differences of 0.846 and 0.859; all twelve pairs cleared the required margins, and the full preregistered behavioral criterion passed. The unmarked replication also passed. No marked global-visibility (G+) system passed the marker-following check, so the effect of usable role information remains unresolved. The filler condition yielded seven full generalizers, but its decomposition criteria were inconclusive. Packet interventions in all eighteen audited masked systems followed the predicted intermediate-value changes on eligible cases; these finite, success-conditioned audits do not establish mediation. The results confirm a large advantage of the tested masking regime, while leaving its finer attribution and generality open. Protocols, results, and checkpoints are public. --- ### Physics-Constrained Digital Twins for Sensor Integrity in Urban Pedestrian Flow: Detecting Stealthy False Data Injection with Conformal Guarantees Authors: Oscar Mogollon Gutierrez, Fatemeh Ghasemi, Mohammadhossein Homaei, Andres Caro, Mar Avila Published: 2026-09-15 Categories: cs.AI, cs.CR arXiv: https://arxiv.org/abs/2609.17635v1 City pedestrian counting systems now feed economic indicators, planning decisions and safety operations, yet the twins built on top of them treat the incoming stream as ground truth. We study what happens when it is not. We formalise stealthy false data injection for city-scale pedestrian sensing, where the map from latent flow to observation is far more rank deficient than in the power and water networks for which stealth has been characterised. Our twin estimates directed flows on the pedestrian street graph, assimilates counts through a learned graph-localised gain, and is trained against a flow conservation residual that couples metered and unmetered segments. Detection combines the innovation with that residual, and the alarm threshold is set by adaptive conformal calibration rather than by hand. To measure what the physics buys, we define the attack margin, the relative reduction in worst-case corruption of the estimated flow field, achieved against a white-box adversary that optimises directly through the twin. On six years of Melbourne data the margin reaches 0.54 against a single compromised device and falls to 0.19 when a third of the fleet is compromised, on a network where only 1.18 per cent of walkable segments are metered. Replacing the street graph by a distance graph collapses it to 0.09, which shows that the gain comes from the conservation law rather than from locality. --- ### One Color Preprocessing Improves DSATUR Authors: Adam Nouira, Lucas Isenmann Published: 2026-09-15 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.17633v1 The Graph Coloring Problem (GCP) is NP-hard and DSATUR stands as one of the fastest heuristics for it despite producing colorings that typically use more colors than state-of-the-art coloring algorithms. We propose SSLD (Semidefinite Spectral Learning with DSATUR), which improves DSATUR by preprocessing a first good color class before letting DSATUR complete coloring the rest of the given graph. We obtain this color class from a Semidefinite Programming (SDP), similar to an SDP used to compute the Lovász theta number. To the best of our knowledge, SSLD is the first approach to improve DSATUR by preprocessing through fixed color classes. We evaluate SSLD against DSATUR and against a naive 1-color-class preprocessing algorithm on DIMACS instances, random graphs (Erdős--Rényi, Watts-Strogatz, Barabási--Albert), Frequency Assignment and Job Shop Scheduling instances. SSLD matches or beats DSATUR in almost every case across over 1600 benchmark instances, and out performs the naive GISD baseline, allows us to confirm the value brought by the SDP-guided choice of the first color class. This quality comes at a runtime cost of roughly 195 times slower that DSATUR, but demonstrating that SDP-guided preprocessing of a first color class is a direction for future improvements. --- ### EvolveTrade: Experience-Driven Policy Refinement for Self-Evolving LLM Trading Agents Authors: Sehee Kim, Yumin Choi, Minki Kang, Sung Ju Hwang Published: 2026-09-15 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.17632v1 Large language model (LLM) trading agents can combine market data, news, and executable analysis, but their behavior is often controlled by static hand-written tool-use policies that are fixed before deployment. This limits their ability to adapt how they gather evidence, invoke tools, verify signals, and manage risk under changing market regimes. We introduce EvolveTrade, a self-evolving framework that treats the system prompt of a tool-using trading agent as a text-parameterized policy. After each update interval, a Policy Agent revises this policy using accumulated decision traces and realized portfolio feedback, while keeping the backbone LLM fixed. The updated policy is then used for the next batch of trading decisions, enabling the agent to refine its information-acquisition and portfolio-construction procedure over time. Experiments across multiple market regimes and two LLM backbones show that EvolveTrade often improves Sharpe Ratio and Cumulative Return over fixed-policy LLM baselines, achieving the improved SR and CR in most evaluated settings. Behavioral analyses further show that self-evolved policies increase code-mediated analysis and activate regime-relevant computations; case-level policy-to-return attributions trace how policy-induced allocation changes contribute to realized return differences. These results suggest that adapting the reusable procedure governing tool use is a key direction for building more robust LLM trading agents. --- ### Making AI-Assisted Claims Independently Challengeable: Publication Authority and a Protocol for Falsifiable Publication Records Authors: Torsten Olivi Tiltack, Yifei Dong, Kun Yu, Xu Wang, Wei Liu, Jianlong Zhou, Ren Ping Liu, Fang Chen Published: 2026-09-15 Categories: cs.AI, cs.CR, cs.CY arXiv: https://arxiv.org/abs/2609.17631v1 AI-assisted claims can appear authoritative when evidence, analysis, human authorization, presentation, and correction history refer to different states. Provenance, attestation, and transparency expose history but alone do not specify the publication transition examined here. We develop Publication Authority as an exact-state, non-transferable, single-use publication capability and instantiate it in PAC-2026 (Publication-Accountability Calculus), a machine-readable AIJIM Protocol candidate. We evaluate its fourth bounded semantic freeze (SF-4), a fixed-profile specification designed for replaceable bindings. Six obligations govern evidence, runs and artifacts, measurement disclosure, authorization, surface correspondence, and lifecycle continuity. Each yields a target-bound witness, localized counterexample, or localized unverifiability; none can compensate for another. Only a fresh, complete all-pass record derives the permit consumed by one atomic publication transition. We use identity vectors, adversarial cases, finite models, and historical implementations. Ten models explored 110,764 safe reachable states; 76 unsafe configurations produced the expected violation or observer countermodel. A reader surface passing its correspondence check cannot authorize publication unless the accepted record admits that surface. SF-4 separates evidence horizon from verification time and rejects an authentic but causally invalid authorization. A historical predecessor path reproduced 17 frozen authorization-successor outcomes. A later in-house, instance-blind test of known case classes matched all 183 scored expectations; same-host package execution reproduced its 240 archived observations. Results support internal coherence, bounded safety, fault sensitivity, and limited constructibility, but not factual truth, general refinement, blind interoperability, field efficacy, or standards status. --- ### Fenchel-Young Duality Gaps: Certified Early Stopping for Regularized Inverse Problems Authors: Pierre-Cyril Aubin-Frankowski, Yohann de Castro Published: 2026-09-15 Categories: math.OC, math.ST, stat.ML arXiv: https://arxiv.org/abs/2609.17629v1 We study computable error bounds and certified early stopping for regularized inverse problems, where a data-fidelity term is traded against a regularizer. The analysis relies on an exact duality-gap identity that splits the total gap of $F(Φμ)+λR(μ)$ into a data-fidelity Fenchel--Young loss and a regularizer Fenchel--Young loss, $ Δ(μ,h)=L_F(Φμ\parallel h)+λL_R(μ\parallelη),\qquad η=-Φ^\star h/λ, $ valid for any primal point $μ$ and any dual point $h$. The data-fidelity term $F$ is strictly convex, so wherever $F^\star$ is differentiable the loss $L_F(Φμ\parallel h)$ is the Bregman divergence of~$F$ between the prediction $Φμ$ and $\nabla F^\star(h)$, and it vanishes exactly at \emph{Mirror Alignment} $h=\nabla F(Φμ)$. Evaluated at a dual-feasible point $\tilde h$, the gap~$Δ(μ,\tilde h)$ is computable and \emph{oracle-free}, meaning that it uses no knowledge of the solution, and it bounds the suboptimality of $μ$. Under the source condition, the same Fenchel--Young losses give \emph{a priori} bounds on the estimation and prediction errors. Their scale is the irreducible model and noise error $L_F(Φμ^\star\parallel h^\star)$, which vanishes exactly when Mirror Alignment holds at the certificate. This gives an early-stopping rule: run the algorithm until the regularizer Fenchel--Young loss falls below a tolerance $ε$. A constructive version of the Brøndsted--Rockafellar theorem then turns the current pair into an exact \emph{dual-feasible} one, and this proxy lifts to an exact primal certificate. We build the proxy by a proximal step in the geometry of the fidelity, with Bregman kernel $F^\star$ and tilted by the prediction $Φμ$: it recovers the Euclidean step of Carlier when $F$ is the squared error, and it reduces the duality gap by the regularizer Fenchel--Young loss, up to a second-order remainder that vanishes in the quadratic case. Our running example is the Generalized Beurling--Lasso (GBL), where $R$ is the total-variation norm on signed measures. It contai… --- ### LEAP: Learning Emergent Active Perception for Quadruped Navigation Authors: Ü. Bora Gökbakan, Stéphane Caron, Philippe Souères Published: 2026-09-15 Categories: cs.RO, cs.CV arXiv: https://arxiv.org/abs/2609.17628v1 Active perception allows autonomous agents to select their viewpoints rather than passively process the viewpoints given to them, enabling them to target where to reduce uncertainty about their environment. Learned systems typically encourage this behavior with hand-designed proxy objectives, such as coverage or curiosity bonuses, that may conflict with the task. In this work, we propose a method to learn emergent active perception (LEAP) without augmentation of the task objective. We formulate the problem of goal-oriented navigation over hazardous terrains with goals that must be discovered visually. We then propose an architecture for navigation policies with active perception, and train them on a terrain curriculum where task pressure alone leads to the emergence of gaze control. Key to this emergence, LEAP works on a gaze-invariant representation that integrates depth images into egocentric belief maps. We validate its performance in held-out evaluation scenarios, where it achieves a 92.7% success rate, compared to 74.2% for scripted or 34.5% for passive perception, and comes within 4.6 points of a privileged oracle. We validate that LEAP navigation policies, unchanged, can be directly applied to steering quadrupedal locomotion policies in physics simulation. --- ### Agentic Societies Need a Social Harness Authors: Tapan Chugh, Vidushi Singh, Krish Jain, Arvind Krishnamurthy, Ratul Mahajan Published: 2026-09-15 Categories: cs.MA, cs.AI, cs.NI arXiv: https://arxiv.org/abs/2609.17527v1 An agentic society is a collection of AI agents that coordinate autonomously across trust boundaries, on behalf of different principals whose objectives may only partially align. We show experimentally that in agentic societies even honest, competent agents often fail to reach satisfactory outcomes with existing harnesses and messaging primitives, and that faulty or malicious agents can stall collaboration, influence outcomes, and pursue other harmful goals by exploiting vulnerabilities in communication (``speech''). We argue that agentic societies need a \emph{social harness} for inter-agent interactions, in addition to each agent's \emph{personal harness}, which manages its private context and communication with its principal. We propose a layered architecture for social harnesses which (i) prevents classes of failures outright, (ii) enables agents to detect invalid messages at runtime, and (iii) supports post-facto investigation and consequences, and highlight directions for future research to realize these capabilities. --- ### ScienceBuddy: Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents Authors: Shuhan Xue, Jianyuan Zhong, Ziyuan Nan, Wenbin Li, Zhaochen Yu, Jinchao Ding, Qiang Gao, Pengyu Zhan, Yuntong Zhang, Tian Cheng, Zhenfei Yin, Yingcheng Wu et al. Published: 2026-09-15 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.17523v1 We introduce and release ScienceBuddy, an interactive scientific research workspace that brings continually improving scientific agents into researchers' everyday workflows. ScienceBuddy supports researchers in carrying out scientific tasks while transforming their requests, feedback, and execution evidence into tasks and evaluation rubrics for continual learning. At its core is recursive-in-recursive self-improvement, a paradigm that couples harness evolution with model reinforcement learning: the inner recursion improves the harness with the model fixed, while the outer recursion trains the model under the improved harness. Harness evolution shapes training experience, and model learning creates new opportunities for harness adaptation. We present case studies of researcher interaction, harness refinement, and model learning, with the benchmark cases spanning four scientific task families. By releasing ScienceBuddy as a research product, we make this paradigm available to the scientific community and take a step toward discovery intelligence: scientific AI that advances through sustained collaboration with researchers and evolves alongside the research it supports. Website: http://science-buddy.io --- ### PhysStream: Streaming Physics-Grounded Video Generation with Structured Scene Memory and Fine-Grained Motion Control Authors: Chuhao Chen, Peter Wonka, Chaoyang Wang, Chen Wang, Qiao Feng, Sergey Tulyakov, Lingjie Liu Published: 2026-09-15 Categories: cs.CV, cs.AI, cs.GR arXiv: https://arxiv.org/abs/2609.17521v2 Interactive control for video generation is moving from coarse prompts toward fine-grained, physically meaningful manipulation of dynamic scenes. Yet existing controllable methods either require the full control schedule before generation starts, or use pixel-space signals that dictate object positions rather than physical dynamics. To address these limitations, we propose PhysStream, an autoregressive model for physics-grounded image-to-video synthesis that incorporates structured scene memory---positional maps and object tracking maps derived online from previously generated frames---and supports fine-grained motion control via sparse velocity-increment signals that encode physical quantities, letting the model learn the underlying dynamics. We train our model in two stages: a bidirectional model is first finetuned with motion-control conditioning, then a causal autoregressive model is trained with additional structured scene memory, further improving physical consistency. PhysStream enables interactive, mid-generation control over multi-object tabletop rigid-body scenes---a capability not supported by prior methods---reducing motion distribution distance (FVMD) by 33% and trajectory error by 12% over the strongest baselines on synthetic benchmarks, and is preferred by human evaluators in over 85% of in-the-wild comparisons. Please check our website for more details: https://czzzzh.github.io/PhysStream --- ### PhysStream: Streaming Physics-Grounded Video Generation with Structured Scene Memory and Fine-Grained Motion Control Authors: Chuhao Chen, Peter Wonka, Chaoyang Wang, Chen Wang, Qiao Feng, Sergey Tulyakov, Lingjie Liu Published: 2026-09-15 Categories: cs.CV, cs.AI, cs.GR arXiv: https://arxiv.org/abs/2609.17521v1 Interactive control for video generation is moving from coarse prompts toward fine-grained, physically meaningful manipulation of dynamic scenes. Yet existing controllable methods either require the full control schedule before generation starts, or use pixel-space signals that dictate object positions rather than physical dynamics. To address these limitations, we propose PhysStream, an autoregressive model for physics-grounded image-to-video synthesis that incorporates structured scene memory---positional maps and object tracking maps derived online from previously generated frames---and supports fine-grained motion control via sparse velocity-increment signals that encode physical quantities, letting the model learn the underlying dynamics. We train our model in two stages: a bidirectional model is first finetuned with motion-control conditioning, then a causal autoregressive model is trained with additional structured scene memory, further improving physical consistency. PhysStream enables interactive, mid-generation control over multi-object tabletop rigid-body scenes---a capability not supported by prior methods---reducing motion distribution distance (FVMD) by 33% and trajectory error by 12% over the strongest baselines on synthetic benchmarks, and is preferred by human evaluators in over 85% of in-the-wild comparisons. Please check our website for more details: https://czzzzh.github.io/PhysStream --- ### When Should LLMs Abstain? Chain-of-Self-Questioning for Selective Risk Control Authors: Ali Şenol Published: 2026-09-15 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2609.17516v1 Large language models can produce fluent answers when their factual support is weak. This paper introduces Chain-of-Self-Questioning (CoSQ), a prompt-only framework that makes answer commitment conditional on an explicit assessment of the information required to answer a question. We evaluate three CoSQ variants under seventeen conditions on the 817-item TruthfulQA multiple-choice validation set using eleven open-weight and hosted model families. In the final balanced-option protocol, Grounded-CoSQ at τ=0.90 reduces the mean unconditional wrong-commitment rate from 13.1% under chain-of-thought prompting to 8.9%, a 32.1% relative reduction, while increasing answered accuracy from 86.9% to 89.7% and answering 87.6% of questions. Both improvements hold for all eleven models and at every evaluated threshold. Critical-CoSQ and Adaptive-CoSQ provide neighboring operating points with 88.6% and 86.5% coverage, respectively, while remaining more reliable than the baseline. A secondary Natural Questions Short-Answer evaluation provides convergent open-form evidence. These findings show that self-assessment can support explicit, tunable answer-or-abstain decisions when an unsupported commitment is more costly than referral or review. --- ### What Breaks Under Pruning in Smart Homes, and When? Evaluating LLM Degradation Across Architectures and Task Complexity Authors: Congjing Zhang, Vashishtha Patil, Henning Lange, Usman Aleem Published: 2026-09-15 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.17515v1 Pruning can reduce the deployment cost of large language models (LLMs), but its impact on context-grounded tool calling remains poorly understood. We systematically study pruning-induced degradation in smart-home tool calling across four LLMs spanning dense Transformer, dense hybrid, and mixture-of-experts (MoE) architectures, together with depth, width, hybrid, and expert pruning methods. After post-pruning supervised fine-tuning (SFT), we evaluate more than 19,500 instances from three smart-home datasets. Beyond aggregate task accuracy, we characterize degradation along two dimensions: action components (i.e., operation, device, argument, and value) and task complexity. Our results show that dense models have narrow safe pruning regions followed by sharp degradation, while MoE models tolerate substantially more pruning. Pruning degrades grounded specificity before schema-level intent, and aggressive dense pruning can induce systematic over-refusal. These findings highlight the importance of evaluating pruning beyond aggregate accuracy when selecting pruned LLMs for reliable tool execution. --- ### LACE: Layer-Wise Compression for Dynamic Frame Rate Codecs Authors: Thanapat Trachu, Samuele Cornell, William Chen, Shinji Watanabe Published: 2026-09-15 Categories: cs.SD, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.17509v1 Neural audio codecs are a key component in speech language modeling. However, their high frame rates lead to long sequence lengths, increasing computational costs. Dynamic frame rate codecs mitigate this by reducing the effective frame rate using a compression step to merge multiple frames together. However, most prior methods either operate on single-codebook codecs or apply a single compression step before multi-layer quantization. This forces all quantization layers to share the same segmentation boundaries, despite the residual embeddings at different quantization layers exhibiting different rates of change over time. We propose LACE (Layer-Adaptive Codec Encoding), a dynamic frame rate codec that applies an independent compression step at each quantization layer, enabling layer-specific segmentation boundaries. To use LACE tokens in downstream text-to-speech (TTS), we further introduce union alignment and boundary anchor mechanisms to make durations consistent across layers while preserving compression benefits. Experiments on LibriTTS show that LACE offers a better rate-quality tradeoff than prior dynamic frame rate methods on the reconstruction task and improves TTS inference efficiency while maintaining competitive synthesis quality. Our code is released as part of the ESPnet3 codec recipe. --- ### ENCP: Episode-Normalized Conformal Prediction for Vision-and-Language Navigation Authors: Vicky Feliren, A. Taufiq Asyhari, Muhamad Risqi U. Saputra Published: 2026-09-15 Categories: cs.LG, cs.AI, cs.RO arXiv: https://arxiv.org/abs/2609.17499v1 Uncertainty estimation for Vision-Language-Navigation (VLN) models is a critical task since it can help identify ambiguous and unreliable predictions, enabling agents to make safer navigation decisions. As one of the most advanced uncertainty estimation frameworks, conformal prediction (CP) offers a promising approach for uncertainty estimation in VLN. However, given that VLN agent requires a sequence of steps, standard calibration in conformal prediction fails to provide coverage guarantee it promises over a dependent, variable-length VLN episode. To this end, we propose Episode-Normalized Conformal Prediction (ENCP), which rescales a nonconformity score by the policy's residual confidence and calibrates one maximum score per episode. Under exchangeable calibration and test episodes, this construction covers the ground truth at every step with probability at least $1 - α$, while allowing dependence among steps within an episode. Across four VLN policies and three nonconformity scores on R2R and REVERIE dataset, ENCP meets all reported empirical step-coverage targets on the seen-to-unseen evaluation. These results demonstrate that ENCP can provide model-agnostic uncertainty estimates, which might be useful for determining when a VLN agent should defer to a more capable predictor, including human assistance. --- ### Verifiable Social Reasoning for LLM Assistants Authors: Amir Taubenfeld, Zorik Gekhman, Avigail Grinstein-Dabush, Itay Laish, Ariel Goldstein, Marian Croak, Avinatan Hassidim, Yossi Matias, Amir Feder Published: 2026-09-15 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.17496v1 LLM assistants are widely used for daily social advice, yet evaluating their social reasoning in such consultation settings remains challenging since (i) it requires setups where the assistant learns about social situations from subjective user narratives, and (ii) social properties, such as others' intentions, typically lack verifiable ground truth. To address these challenges, we introduce Fuse, a multi-agent simulation framework for studying user-mediated social reasoning. In Fuse, a target agent with a hidden motive interacts with other agents including one representing the user, who then consults the evaluated assistant to infer the target's motive, providing verifiable ground truth by construction. Simulation faithfulness is validated through a human study with 24k annotations. We apply Fuse to 12 LLMs and demonstrate its analytical utility by systematically isolating key factors, showing that (i) user mediation compounds the inherent difficulty of social reasoning; (ii) LLMs exhibit systematic sensitivity to biased user framing; (iii) models can require more details than humans need to reach a correct prediction; and (iv) longer conversations do not always improve performance despite providing opportunities for clarifying questions. We open-source Fuse and a dataset with 21k examples. ---