# Research Papers — 2026-09 · bundle page 182 of 266 > Full-text markdown of papers #9051–9100 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/182 [← Previous bundle](https://aipod.dev/papers/archive/2026-09/page/181.md) [Next bundle →](https://aipod.dev/papers/archive/2026-09/page/183.md) ### Harmfulness Propagation Dynamics: Layer-wise Trajectories of Adversarial Intent in Large Language Models Authors: Noor Islam S. Mohammad, Uluğ Bayazıt Published: 2026-09-11 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.13534v1 We identify \textbf{Harmfulness Propagation Dynamics (HPD)}: for harmful prompts, the projection of the last-token hidden state onto a learned harm direction rises monotonically with transformer depth, whereas benign prompts remain flat or oscillatory. This cross-layer signature reflects harmful intent as a \emph{progressively resolved} semantic property: surface form appears early, while pragmatic intent consolidates later, making the \emph{trajectory shape} more informative than any single-layer snapshot. Moreover, LDA-based harm directions, learned per layer, remain stable across random splits (pairwise cosine similarity $>0.97$), supporting the projection sequence as a reproducible structured signal. Building on HPD, we introduce \textbf{\herald{}} (\textbf{H}armful \textbf{E}ncoding \textbf{R}ecognition via \textbf{A}ctivation \textbf{L}ayer \textbf{D}ynamics). This lightweight input moderator extracts a seven-dimensional feature record, slope, curvature, monotonicity, onset layer, and related statistics from the cross-layer projection sequence and classifies it with a 288-parameter MLP. \herald{} stores one $d$-dimensional direction per layer ($262$\,KB for a 32-layer, $d{=}4096$ model), requires no gradient computation during training, and adds only $2.6{\times}10^{-6}$ prefill FLOPs at inference. Across eight prompt-harmfulness benchmarks and four model families, \herald{} achieves an average F1 of $89.3$ on OLMo2-7B, surpassing all tested guard models on adversarial jailbreak detection ($98.4$ vs.\ $96.9$ F1) and outperforming prior latent-based methods by $2.3$-$4.1$ F1 points on every backbone. Per-instance trajectories provide machine-readable audit records that reveal \emph{when} and \emph{how} harmfulness emerges, offering an interpretability advantage over single-layer approaches. --- ### Early-Stopping Thresholds for ES-HyperNEAT: A Data-Driven Approach from Fitness Dynamics Authors: Romain Claret, Arthur Gygax, Michael O'Neill, Paul Cotofrei, Pascal Felber Published: 2026-09-11 Categories: cs.NE, cs.LG arXiv: https://arxiv.org/abs/2609.13533v1 Most hyperparameter configurations for Evolvable-Substrate HyperNEAT (ES-HyperNEAT) produce networks that stagnate at random-guessing performance, wasting computational resources. We frame early stopping as binary classification on early fitness trajectories: for each trial, we compute the cumulative median of best-per-generation fitness and test it against a threshold derived by maximizing the F1 score on an initial 90-trial dataset. The resulting rule (generation G* = 3, threshold T* = 0.140) achieves F1 = 0.872 on 180 independent validation trials, retaining over 90% of successful trials while cutting computational cost by 41.6%. Compared to Hyperband, our domain-specific rule is 64% more efficient with higher mean fitness, though Hyperband occasionally discovers higher peak solutions. On a converged search population the rule becomes too aggressive (recall 31.1%), motivating adaptive thresholds. The specific thresholds are ES-HyperNEAT-specific, but the methodology, deriving stopping criteria from fitness dynamics classification, is applicable to other evolutionary algorithms with stagnation-prone hyperparameter spaces. --- ### Attention Is All You Need (to Avoid Spurious Oscillations) Authors: Jinyoung Jeong, Joseph B. Choi, Xinlun Cheng, H. S. Udaykumar, Sanghun Choi, Stephen S. Baek Published: 2026-09-11 Categories: cs.LG, cs.AI, physics.flu-dyn arXiv: https://arxiv.org/abs/2609.13531v1 Can attention move a shock across several cells in one update without breaking it? We develop a conservative, fixed grid finite-volume scheme in which a CFL-conditioned attention flux selects upstream information according to the transport required by the current time step. One-dimensional inviscid Burgers transport is used as the central mechanism test: the same learned flux remains reliable in the conventional small-step regime and, with a time step four times larger, preserves sharp shocks while using one stage per update. A standard fifth-order WENO scheme with third-order strong-stability-preserving Runge-Kutta time integration (WENO-5+SSP-RK3) is included alongside controlled Forward Euler comparisons to separate flux selection from time integration. The learned attention shifts upstream with the local transport reach and becomes more selective near shocks; inference-time interventions and retrained ablations show that transport-scale information and state-dependent selection contribute directly to performance. Directional two-dimensional scalar Burgers transport and the one-dimensional shallow-water system then test whether the conservation-scale-selection principle transfers beyond the original scalar setting. The results support attention as a learnable information stencil for conservative large-step shock transport, while identifying finite candidate reach and problem-dependent robustness as the present limits. --- ### Generative Interpretability via Scalable Neuro-Symbolic Models Authors: Xiaocong Yang Published: 2026-09-11 Categories: cs.LG, cs.AI, cs.CL, cs.SC arXiv: https://arxiv.org/abs/2609.13529v1 As the use of Large Language Models moves from chatbots into agentic systems, where outputs become actions with irreversible consequences on reality, the existing paradigm on AI Interpretability research, post-hoc interpretability, is structurally inadequate for safe and trustworthy model deployment: it explains behavior after the fact but cannot audit or intervene in an inference computation before it commits to an output. We therefore argue for a shift toward \emph{generative interpretability}, an architectural property under which a model's inference pass natively exposes semantically meaningful checkpoints that are human-understandable and amenable to causal intervention. We show the merits of generative interpretability as comparison to other interpretability research paradigms, and propose Neuro-Symbolic Models as a concrete instantiation. --- ### A family of spectral conjugate gradient algorithms derived by least-squares approximations based on a modified quasi--Newton update with application to a revised robust binary classification model Authors: Saman Babaie-Kafaki, Maryam Khoshsimaye-Bargard, Ahmad Mousavi Published: 2026-09-11 Categories: math.OC, cs.LG arXiv: https://arxiv.org/abs/2609.13526v1 We develop a spectral three-term modification of the classic Hestenes--Stiefel conjugate gradient algorithm, preserving its anti-jamming characteristic and, simultaneously, taking care of the sufficient descent property. We discuss how a modified secant equation can be extracted from our modification scheme, yielding a memoryless BFGS updating formula. Then, the spectral parameter of our method is obtained by steering its direction toward the given BFGS direction within a least-squares context. Using our technical improvements, we outline the general framework of our algorithm and discuss its theoretical features, including the descent and convergence properties, without the convexity assumption. We put our algorithm to the test in comparison with the three other conjugate gradient algorithms on a set of CUTEr unconstrained optimization test models, comparing the outputs using the Dolan--More measure. Next, we provide a concise evaluation of the results, highlighting the practical advantages of our algorithm. As a real-world case study, we introduce a reduced quadratic surface SVM with the rescaled loss for robust binary classification and apply the proposed algorithm to assess its accuracy and training time against several other SVMs. --- ### Rolling Day-Wise Mortality Prediction in Critically Ill Patients With AKI on CRRT Utilizing Machine Pressure Waveforms Authors: Shehan Irteza Pranto, Joanna Yang, Joshua Lambert, Stuart L. Goldstein, Lili Chan, Girish N. Nadkarni, Tiago K. Colicchio, Javier A. Neyra, Jin Chen Published: 2026-09-11 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.13524v1 Critically ill patients with acute kidney injury (AKI) on continuous renal replacement therapy (CRRT) face high mortality, yet current risk assessment relies primarily on clinical parameters from electronic health records (EHR) and ignores minute-level circuit pressure waveforms generated by CRRT machines that track the extracorporeal circuit's interaction with the patient. Clinicians therefore cannot see deterioration as it develops. Risk is reassessed only when labs are drawn, while this continuous record is discarded because it is contaminated by shared-device records, non-physiological minutes, and sensor artifacts. To make the stream usable, we aligned machine records to charted therapy intervals to prevent cross-patient leakage, removed priming and downtime minutes, tuned denoising on a synthetic spike-injection benchmark, and masked unobserved intervals rather than imputing them. On this cleaned stream, we define a rolling day-wise task and a transformer-based stacked ensemble that late-fuses a window-reduced sequence transformer with classical models using circuit-instability features and clinical EHR variables. In a leak-safe benchmark on the multi-center CRRTnet cohort (976 patients, 4,585 treatment days), the machine-only model had the lowest standalone prognostic value (AUROC 0.625), followed by the EHR-only model (0.717). Integrating EHR and machine streams reached a one-day mortality AUROC of 0.766. SHAP attribution showed that circuit-instability descriptors raised the machine share of the top 15 combined-model features from 3 to 7 (20.0% to 46.7%), highlighting filter pressure, transmembrane pressure (TMP), and access-to-return difference (ARD). To our knowledge, this is the first patient-level mortality prediction incorporating CRRT machine data, turning a discarded bedside stream into a continuous risk signal. --- ### Positive Topology and Feasible Refinement: Forcing Matrices, Positivity, and Information Authors: Mirco A. Mannucci, Giovanni Sambin Published: 2026-09-11 Categories: cs.LO, cs.AI arXiv: https://arxiv.org/abs/2609.13523v1 We develop a conceptual and operational account of Positive Topology starting from a basic relation between points or models and observable properties. From this relation, two complementary structures emerge. The first captures universal refinement and cover: what must hold across all relevant cases and how information can be systematically refined. The second captures positivity and witnessed existence: what can be positively realized and sustained without relying on classical complements. A central result shows that the underlying relation between points and observables can be reconstructed from either of these induced structures. We also clarify the distinction between point-based and pointfree formulations: when points are available, positivity can be derived from the underlying forcing relation, while in the formal pointfree setting positivity is taken as primitive and its compatibility with cover is imposed axiomatically. We develop two complementary interpretations of the framework. The first is information-theoretic, viewing cover as refinement of partial information and positivity as witnessed feasibility. The second is game-theoretic, viewing positivity as the ability of a witness or hypothesis to survive successive legitimate refinements. The final part of the paper is deliberately programmatic. We outline how resource constraints, verification costs, and finite budgets can be incorporated into the framework. This leads to resource-sensitive notions of forcing, cover, positivity, and refinement, and raises new questions about how these structures behave as available resources change. Examples from medical diagnosis, legal reasoning, and AI systems illustrate the potential relevance of the approach to grounded, explainable, and resource-aware inference. --- ### From Token Probabilities to Semantic Constraints: Towards Declarative Probabilistic Evaluation of Language Models Authors: Kyle Richardson, Cullen Anderson, Pranav Balakrishnan, Takuto Ban, Daksha Ladia, Ankita Gupta, Marisa Hudspeth Published: 2026-09-11 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.13520v1 While Large Language Models have improved rapidly, many fundamental questions remain about how to evaluate the knowledge and reasoning abilities they acquire, and how such evaluations relate to the learning signals used in pre-training. In this paper, we propose ModelLog, a declarative probabilistic framework for pre-training evaluation that makes the semantic structure of model behavior explicit and provides new formal tools for relating evaluation to learning. ModelLog specifies evaluation targets as symbolic constraints over token-level predictions and measures how strongly a model's distribution satisfies those constraints. We explore the framework through a new suite of tasks targeting negation, mutual exclusivity, and consistency, finding systematic failures that are difficult to characterize through token likelihood or answer accuracy alone. We further show that these evaluation scores can also be interpreted as losses, whose gradients reflect logical strength, informativeness, and variable-level sensitivity. This links evaluation and learning through a shared semantics, suggesting evaluation methods that diagnose model behavior while also helping to clarify the semantic structure of learning. --- ### Fraglingo: Molecular Design via Attachment-Aware Autoregressive Fragment Generation Authors: Thao Nguyen, Jeonghwan Kim, Zhenhailong Wang, Heng Ji Published: 2026-09-11 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.13519v2 We introduce Fraglingo, an autoregressive molecular generator that constructs molecules step by step from chemically meaningful fragments connected through predefined attachment sites. At each generation step, Fraglingo jointly predicts which fragment to add and how it should attach by producing an attachment-aware fragment embedding and retrieving the nearest fragment through latent-space search. A wildcard-anchored readout represents both the growing molecule and candidate fragments relative to their attachment sites, enabling a single latent prediction to determine both fragment identity and attachment configuration. Because prediction operates in a continuous embedding space rather than over fixed fragment identifiers, larger fragment libraries can be introduced at inference time without retraining. This retrieval-based formulation provides a unified generation primitive for molecule generation, scaffold generation, scaffold decoration, and molecule optimization. Fraglingo also supports property-conditional generation, allowing desired molecular properties to guide the generation process. On controlled property-conditional benchmarks, Fraglingo achieves stronger joint property control than comparably trained baselines while maintaining competitive validity, uniqueness, and novelty. It also generalizes to fragment libraries up to four times larger than those used during training without retraining. Code is available at: https://anonymous.4open.science/r/FragLingo-3551. --- ### GeoTTER: Leveraging Local Geometry of Optimal Transport for Zero-Shot Classification Authors: Wei-Yang Alex Lee, Rudrasis Chakraborty, Vishnu Lokhande Published: 2026-09-11 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.13518v1 We present GeoTTER, a novel framework that redefines optimal transport in the realm of zero-shot classification. Conventional methods often suffer from miscalibration and a lack of adaptability, as they rely on fixed cost matrices derived solely from pre-trained model embeddings. In contrast, GeoTTER addresses these limitations by incorporating two key techniques. First, to alleviate high-frequency label jaggedness (sample-level manifold jitter that assigns neighboring embeddings to different classes), GeoTTER integrates local geometric structure into the optimal transport formulation via graph-Laplacian smoothing, a technique grounded in spectral graph theory that enforces neighborhood consistency. Second, to correct coherent angular drift (a low-frequency orientation bias in which large groups of samples share the same angular offset from their true label prototypes), we fuse clustering-guided cost components with a globally adjusted transport cost, achieving a multi-objective optimization that respects both global distribution constraints and latent data structure. With a median improvement of +6.82% compared to zero-shot and +2.13% compared to OTTER, GeoTTER shows robust improvements across a diverse set of benchmarks. --- ### Operational Range Bounding in Spectroscopy: A Safety Cage Framework for Machine Learning Models Authors: Nikki Grens, Luís F. Simões, Kai Hou Yip, Theresa Lueftinger Published: 2026-09-11 Categories: cs.LG, astro-ph.EP, astro-ph.IM arXiv: https://arxiv.org/abs/2609.13514v1 Ensuring the reliability of black-box machine learning models in safety-critical space missions remains a significant challenge, particularly when ground-truth is unavailable for validation. Although machine learning models offer a powerful means to augment standard pipelines by extracting transmission spectra from complex exoplanetary light curves, their susceptibility to unmodelled instrument anomalies, stellar activity, and domain shifts introduces unquantified risks. This study evaluates a modular safety cage architecture that operates as a parallel monitoring layer to assess the validity of a prediction without modifying the underlying estimator. By monitoring different runtime indicators, including uncertainty quantification, out-of-domain detection, and influence functions, the framework constrains the model's operational domain to a verified region. A controlled evaluation is conducted under both in-domain and cross-domain conditions, using datasets from the 2019 and 2021 editions of the Ariel Data Challenges. The results reveal that model failure is multifaceted and that no single indicator captures all failure modes, demonstrating the need for indicator fusion. The application of safety-driven rejection strategies shows that a modest 20% reduction in data coverage results in error reductions between 45% and 65% across different domains and evaluation metrics. Using a formalised coverage-risk framework, a systematic analysis of indicator combinations is performed to identify configurations that maximise risk-ranking accuracy and optimise the trade-off between data coverage and scientific performance. Safety cages provide a transparent mechanism for detecting unreliable predictions and represent a critical step towards the safe deployment of data-driven models in scientific applications, such as astrophysics, where ground truth is seldom available. --- ### Adaptive Phase-Switching for Communication-Efficient Federated LoRA Fine-Tuning Authors: Jerry Adams Franklin Published: 2026-09-11 Categories: cs.LG, cs.AI, cs.DC arXiv: https://arxiv.org/abs/2609.13512v1 Federated fine-tuning of large language models with low-rank adaptation reduces per-client trainable parameters, but client-to-server communication remains the dominant cost. Existing accounting for federated LoRA protocols omits the asymmetric transition round when a protocol changes aggregation mode, and reports savings that ignore grouped-query attention shapes. This paper measures per-round upload and download bytes for a bidirectional B-only federated LoRA protocol and places five methods, three from prior work, on a single communication-quality frontier. The frontier has a knee, which an adaptive phase-switching aggregator, ReverseAdaptive, locates by monitoring the relative improvement in global training loss against a dimensionless threshold rather than by fixing a phase boundary in advance. On TinyLlama-1.1B-Chat with Alpaca, ReverseAdaptive attains 40.5 percent measured round-trip savings over FLoRA at a held-out instruction-following loss cost of 0.0063. It outperforms FFA-LoRA, which freezes the first of the two LoRA factors at initialization, by 0.0182 in held-out loss, more than twenty times the largest per-method seed standard deviation on that metric, so learning that factor before freezing it produces better adapters. The same threshold transfers across model scales without retuning, and the quality cost of the transition is stable across the two datasets tested. --- ### A Three-Axis Stress Test of LLM vs Classical ML for Network Intrusion Detection under Distribution Shift and Adversarial Evasion Authors: Muhammad Ebad Atif, Muhammad Haider Ali Published: 2026-09-11 Categories: cs.LG, cs.AI, cs.CR arXiv: https://arxiv.org/abs/2609.13511v1 Large language models are increasingly benchmarked against classical machine learning for network intrusion detection (NIDS), almost always using same-dataset evaluation, and that protocol turns out to be incomplete. Evaluating XGBoost and RoBERTa-LoRA on two independently collected NetFlow v2 networks across three axes (same-dataset performance, cross-dataset transfer, and adversarial evasion) reveals no universal winner. The two models are statistically tied same-dataset. XGBoost wins decisively under cross-dataset distribution shift, by 15 points of F1 and 25 points of balanced accuracy; on the target network RoBERTa-LoRA's false positive rate reaches 0.78, leaving it barely above chance despite a superficially moderate F1. RoBERTa-LoRA wins decisively under adversarial evasion, by roughly 17 points of F1 at a representative mid-range perturbation strength, while both models hold false positive rates below 0.01 throughout. The model an evaluator would recommend therefore depends entirely on which axis is tested, not on same-dataset accuracy alone. A staged feature-leakage ablation improves cross-dataset transfer non-monotonically, indicating the leakage signal is distributed across the feature representation rather than confined to a few columns, and cross-dataset transfer between our two networks is strongly directional. These results argue for evaluating NIDS models along multiple independent robustness axes, and with more than one metric per axis. --- ### Pretraining for Sample-Efficient Neural Interfaces Authors: Ben Tang, Zachary Spalding, Gregory B. Cogan Published: 2026-09-11 Categories: cs.LG, q-bio.NC arXiv: https://arxiv.org/abs/2609.13507v1 Brain-computer interfaces (BCIs) decode neural activity to restore lost function. Typically, training a high-performance neural decoder requires a large labeled dataset to be collected from every new subject. One way to reduce the labeled data cost is self-supervised pretraining, which learns general neural representations from unlabeled recordings that accumulate across subjects. However, for intracranial electroencephalography (iEEG) recordings, self-supervised learning has been challenging due to differences in contact placement and neuroanatomy between subjects. We propose MAPA, an otherwise vanilla masked autoencoder with two spatial encodings, an anatomical region embedding and a relative positional encoding, which together enable it to learn neural representations that transfer to unseen subjects and across various tasks. MAPA sets a new state of the art across all three regimes of the Neuroprobe benchmark without fine-tuning: within-session, cross-session, and cross-subject. In the cross-subject regime, a linear probe on MAPA's features needs only ${\sim}164$ labeled trials to reach the accuracy that takes 3,500 without pretraining. Our results show that self-supervised pretraining can scale across heterogeneous iEEG recordings and reduce the labeled data needed for accurate decoding in new subjects. --- ### nBMS, a Neuromorphic Battery Management System with a Silicon-Validated Spiking State-of-Charge Core for eVTOL Aircraft Authors: İsmail Can Dikmen Published: 2026-09-11 Categories: cs.NE arXiv: https://arxiv.org/abs/2609.13506v1 State-of-charge (SoC) estimation for electric vertical take-off and landing (eVTOL) aircraft must run on the vehicle under hard energy and certification budgets, on a duty cycle unlike anything in the automotive literature. In this study an event-driven spiking network, the state-estimation core of the nBMS neuromorphic battery management architecture, is designed for per-timestep SoC estimation and evaluated on a public 22-cell eVTOL dataset with an automotive cross-check. A delta and population encoder, a second-order sigma-delta spiking layer, and a rate-accumulator readout hold 34{,}433 int16 parameters with zero dense multiply-accumulate (MAC) operations. The estimator reaches 2.45\% root-mean-square error (RMSE) against 1.74\% for a tuned long short-term memory (LSTM) baseline, a gap characterized as a temporal-mixing limit in cruise; in exchange it degrades 1.6 times slower than an adaptive leaky integrate-and-fire control under sensor noise and needs roughly 3{,}500 additions per step where the LSTM needs 67{,}700 multiply-accumulates. The full core is deployed on a low-cost automotive-qualified Artix-7 field-programmable gate array at 87\% block-RAM utilization, meets timing at 50~MHz, and reproduces the frozen fixed-point reference bit-exactly over a real 491-step flight segment on silicon. Fully annotated post-implementation analysis gives 1.30~$μ$J per inference, an average of 0.65~$μ$W at the 0.5~Hz mission cadence. --- ### RIGOR: Rig-Informed Geometry for Omnidirectional Reconstruction Authors: Tingjun Huang, Dmitry Rudshin, Mathieu Meyer, Pietro Bonazzi, Marc Pollefeys, Emilia Szymańska Published: 2026-09-11 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.13504v1 Recent developments in feed-forward 3D reconstruction resulted in models which can recover dense scene representations and camera motion solely from an image stream. However, such predictions are prone to becoming inconsistent over long trajectories, specifically in demanding environments with repetitive structures, weak textures and dynamic objects or people. One way to mitigate those challenges is to use an omnidirectional camera, which provides wide spatial coverage and captures richer visual information. Yet, the majority of models do not offer support for 360-degree imagery or require additional fine-tuning. To bridge these two aspects, we present RIGOR: a large-scale reconstruction pipeline for gravity-aligned omnidirectional videos that retains a frozen feed-forward perspective backbone and exploits each panorama as a four-view virtual rig. The rig structure is used to detect and repair locally inconsistent predictions, to retrieve loop closures through cyclic four-view consensus, and to geometrically verify candidate revisits before global optimization. Verified constraints drive a Sim(3) pose graph that corrects accumulated rotation, translation, and scale drift along the sequence. We demonstrate that the proposed consistency mechanisms improve both trajectory accuracy and reconstructed geometry over a feed-forward baseline on challenging construction-site sequences. The code is made available under this link: https://github.com/TangentH/RIGOR. --- ### One Spectrum, Two Resources: Data-Memory Scaling in Autoregressive Prediction Authors: Chiwun Yang, Xiaoyu Li Published: 2026-09-11 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.13500v1 How much learned memory is needed to benefit from more data? We show that the two resources are governed by one predictive-energy spectrum in a positive-entropy autoregressive retrieval source. Each coordinate contributes its query probability times the squared radius of its unknown logit. Writing $μ$ for the resulting energy spectrum, we prove the minimax law $\mathfrak R^*_{\rm value}(n,B)\asymp_R Φ_μ(n^{-1})+Φ_μ(τ_B), Φ_μ(t)=\int\min\{x,t\}\,μ(\mathrm dx),$ for $n$ prediction blocks and a learned state with at most $2^B$ values. Data set the resolution $1/n$; memory sets the level $τ_B$ reached by optimal bit allocation. The complete curve also recovers the positive spectrum. Energy-dimension pairing is essential: two causal sources with identical block-energy and block-dimension marginals have different data and memory exponents. A masked query-key attention head learns the route and values, realizing the law with explicit routing, format, and arithmetic errors. Further results give exponent-adaptive allocation, finite-precision realization, and compute-precision laws under two-sided arithmetic assumptions. Experiments recover the data-memory collapse and coupling exponents, explain the routing and allocation mechanisms, and examine weight-only quantization across six pretrained-model scales. --- ### Canaries in the Bank: Auditing User-Level Privacy in Private Evolution Authors: Sai Aparna Aketi, Enayat Ullah, Shripad Gade Published: 2026-09-11 Categories: cs.CR, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.13499v1 Private Evolution (PE) generates high-fidelity synthetic data in federated settings without exposing users' raw data. It aggregates clipped user votes over a shared candidate bank into a differentially private histogram, with noise calibrated to the worst-case user contribution. However, it is unclear whether an adversary can realize this worst-case privacy loss while following the PE protocol. We introduce a protocol-aware empirical audit in which the server commits to a single shared candidate bank and replaces roughly 1% of its entries with probes derived from a known, non-private canary. We evaluate eight attacks, including an unchanged-bank baseline, exact copies, plausible paraphrases, and high-entropy synthetic nonces. Experiments on Yelp and Sentiment140 show that natural-text attacks remain substantially below the theoretical DP bound, while nonce-based attacks yield considerably stronger bounds and come closest to the mechanism's privacy ceiling. These results quantify the gap between formal worst-case privacy and leakage achievable through protocol-valid candidate-bank manipulation. --- ### Building a Production Greek-English Speech Recognizer Authors: Christos Petrocheilos, Cleopatra Papadopoulou, Chris Porikis, Ioakeim Perros, Ayoub Kirouane, Themistoklis Nikolis Published: 2026-09-11 Categories: eess.AS, cs.AI, cs.SD arXiv: https://arxiv.org/abs/2609.13498v1 We report a multi-month engineering program to build Sophea, a production bilingual Greek-English automatic speech recognition system. We evaluate the system against nine production gates covering Greek and English word error rate, language identification, and hallucinations on non-speech audio. Across twenty-three training iterations and two model architectures, no training-data composition passed all nine gates simultaneously. Meeting the Greek noisy-environment target required about 1,500 steps of dense domain exposure, while preserving English language identification tolerated only about 250 steps, or about 1,250 with a rebalanced mix that reduced Greek accuracy. We describe a six-stage data pipeline in which calibrating an audio-quality filter against in-domain anchors reduced the discarded share of scored Greek audio from 98.7 percent to 10.6 percent. A pre-registered ablation isolated a hallucination defect to one training-data package. A three-model ROVER ensemble increased gate coverage from 4-7 of 9 for individual models to 9 of 9 and reduced overlapping-speech WER from 53.35 percent to 37.87 percent, a 29 percent relative improvement. A separate learned per-clip arbiter over two models is listed as sophea/asr-k1 (preview) on the public Open ASR Leaderboard, with 4.26 percent average WER across eight public English test sets, and reaches 25.88 percent WER on live Greek noisy-environment traffic. We also document five cases in which a measurement tool produced a plausible but incorrect result and seven substantial approaches that were evaluated but not shipped. No model weights or training data are released; we report methodology and quantitative results only. --- ### Generative AI and Extended Reality in Collaborative Architectural Design Education: An Exploratory Studio Study Authors: Yao Xiao, Max Chen, Yichen Li, Nathaniel Powers, Maxwell Wiesenfeld, Gillian Smith, Soroush Farzin, Shichao Liu Published: 2026-09-11 Categories: cs.HC, cs.AI arXiv: https://arxiv.org/abs/2609.13494v1 Architectural design education relies heavily on visual ideation and representation to support collaborative learning in studio environments. Recent advances in generative artificial intelligence (GenAI) and extended reality (XR) offer new opportunities for rapid idea exploration and immersive spatial visualization. This exploratory mixed-methods classroom study investigated how GenAI-assisted multi-user XR influenced collaborative architectural conceptual design. We developed GenARch, a pipeline that integrates GenAI-based visual generation with collaborative XR environments, and deployed it in an undergraduate architectural design studio. Twenty-seven students formed seven self-selected design teams; four teams incorporated GenARch into their usual course workflow to support collaborative ideation and visualization, while three teams continued the same course workflow without GenARch. Pre- and post-intervention surveys assessed design self-efficacy, attitudes toward collaborative learning, and teamwork; a seven-member panel evaluated team design presentations; and GenARch teams participated in group interviews. The quantitative results showed larger relative declines in confidence and outcome expectancy for the GenARch condition and a positive difference-in-differences estimate for perceived conflict management, while panel-rated presentation outcomes were not significantly different between conditions. Interviews indicated complementary roles for the technologies: GenAI supported idea externalization and visual reference generation, whereas XR supported spatial, contextual, and scale-based evaluation. Students also reported challenges related to control, dimensional fidelity, shared attention, and motion comfort. These findings highlight both opportunities and limitations when GenAI and XR are incorporated into collaborative design education. --- ### Token Efficient Task Execution via Application Behavior Modeling for Web Agents Authors: Alexandru Ianta, Eleni Stroulia Published: 2026-09-11 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.13491v1 The strong performance of AI Agents across an impressive variety of tasks is driving an unprecedented investment in agentic infrastructures, however the cost of processing tokens is fast increasing. Web agents automate the execution of web-application tasks described in natural language, by analyzing the web-application's user interface (UI) and interacting with it. This work introduces OdoBot, a novel web-agent architecture that completes tasks at a fraction of the cost when compared to conventional web agents. This is achieved by leveraging a behavioral model of the underlying application constructed by analyzing successful task-execution demonstrations. Our experiments with 45 tasks on the Canvas Learning Management System (LMS) demonstrate that OdoBot uses 44% and 80% fewer tokens than two state-of-the-art competitor agents (Agent-E and WebVoyager), while also surpassing WebVoyager in terms of task success rate. --- ### Mixture-of-Experts Language Models Can Be Strong and Efficient Retrievers Authors: Anubhav Shrestha, Safal Shrestha, Minwu Kim, Torsten Suel, Keith Ross Published: 2026-09-11 Categories: cs.IR, cs.CL arXiv: https://arxiv.org/abs/2609.13486v1 Recent work has shown that fine-tuning decoder-only large language models (LLMs) for retrieval yields strong first-stage retrievers, with effectiveness improving as backbones grow in size. However, every query and document must pass through the full model, so encoding cost increases with model size. Mixture-of-Experts (MoE) LLMs activate only a subset of parameters per token and are widely used to scale generative models, yet remain underexplored as retrievers. We systematically study MoE backbones for retrieval by training MoE and dense LLMs from several families using the same procedure, evaluating them across diverse datasets, and measuring query encoding time under the same serving configuration. We show that MoE retrievers outperform dense retrievers with comparable active parameter counts by up to 3.0 nDCG@10 points on BEIR. One of our strongest MoE retrievers matches an 8B dense retriever with 59% fewer active parameters and 18% lower query encoding time. We further show that the number of experts used for query encoding can be reduced without retraining or re-indexing, retaining more than 99% of retrieval effectiveness while reducing query encoding time by up to 26%. Recent rerankers provide only modest additional gains over strong MoE first stages, which often match or exceed the reranked configurations we evaluate. Together, these results show that MoE LLMs can be strong and efficient first-stage retrievers. --- ### A Hybrid Hierarchical 1D-CNN-BiLSTM Framework for Extractive Summarization of Biomedical and Clinical Text Authors: Saad Bin Ather, Muhammad Saif, Ali Hassan Khan, Manzer Abbas, Hajra Waheed Published: 2026-09-11 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.13481v1 Large language models have made abstractive summarization remarkably fluent, but generated summaries can hallucinate facts, posing serious risks in biomedical and clinical domains. We address this by removing generation from the pipeline and framing summarization as extractive sentence selection. Our Hybrid Hierarchical CNN-LSTM Summarizer uses stacked multi-kernel convolutions to compose sentence-level embeddings into richer inter-sentence representations, followed by a bidirectional LSTM to model long-range dependencies across the document. A lightweight scoring head assigns per-sentence importance scores and is trained end-to-end with binary cross-entropy against oracle extractive labels. At inference, a dynamic mean-plus-standard-deviation threshold with a top-3 fallback selects sentences directly from the source and chronologically reorders them into the final summary. Since every output sentence is copied from the input, the model avoids generation-induced factual drift. On PubMed, our architecture outperforms isolated CNN and LSTM baselines, while ablations show that wider convolutional receptive fields improve sentence scoring. On MIMIC-CXR and MIMIC-IV BHC, the model performs well on unstructured narratives but defaults toward positional baselines on highly templated reports. These results suggest that structural constraints can provide a reliable path toward factually grounded summarization systems that are trustworthy by design rather than by correction. --- ### On the Potential of Multi-Task Learning in Predictive Process Monitoring Authors: Lukas Kirchdorfer, Keyvan Amiri Elyasi, Heiner Stuckenschmidt Published: 2026-09-11 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.13477v1 Predictive Process Monitoring (PPM) forecasts how ongoing organizational processes unfold, enabling information systems to move beyond execution support toward proactive analysis and monitoring. Although deep learning has improved prediction accuracy in PPM, most approaches follow a single-task learning (STL) setup, training a separate model per task. This increases maintenance effort and overlooks potential synergies. Multi-task learning (MTL), which jointly learns multiple prediction targets in one model, offers a promising alternative, yet its effectiveness in PPM remains underexplored. It remains unclear whether and under which settings MTL improves upon STL, which prediction tasks benefit most from joint learning, which task combinations are particularly synergistic, and if and how tasks should be balanced. To fill this gap, we present the first comprehensive empirical study of MTL for PPM, evaluating a variety of task combinations, neural architectures, and optimization methods. Overall, our results position MTL as a strong paradigm for PPM: we see substantial improvements in next-activity prediction and inherent mitigation of class imbalance using MTL, while task balancing is especially critical under low-capacity models. --- ### Edge-addition monotonicity of positive p-energy fails for every p >= 1 Authors: Koyar Afrasyab Published: 2026-09-11 Categories: math.CO, cs.AI arXiv: https://arxiv.org/abs/2609.13476v1 At a 2021 AIM workshop, Guo conjectured that the positive square energy s+ = E+_2 should inherit the familiar edge-addition monotonicity of the spectral radius, rho(G + uv) >= rho(G). That conjecture was subsequently shown to fail at p = 2. Tang, Liu, and Wang then introduced positive p-energy, proved nonmonotonicity for every 1 <= p < 3, and in version 3 of their preprint (26 March 2025) explicitly conjectured that monotonicity should hold for p >= 3. We disprove this conjectured high-exponent extension completely: for every real p > 2 there are infinitely many connected graphs G and nonedges uv such that E+_p(G + uv) < E+_p(G). Together with the Tang-Liu-Wang counterexamples below 3, this shows that no exponent p >= 1 restores the spectral-radius-style monotonicity: positive p-energy can decrease under the addition of an edge for every real p >= 1. The construction is a chain of clique blocks joined by regular bipartite graphs. An equitable quotient converges to Q = I + cA(P_k). For noninteger p, a binomial-series sign argument for a fractional power of I - cA(P_k) gives the required negative endpoint entry. At integer exponents, choosing c across the first spectral threshold leaves exactly one negative eigenvalue, and path locality forces the positive spectral contribution to have negative sign. We also give a fully rational 38-vertex certificate at p = 4 and determine the complete failure interval of a fixed 17-vertex counterexample at p = 3. --- ### Grounded Adjudication of Variations across Extracted TimeLines (GAVEL): Comparing Clinical Timelines Against Their Case Reports Authors: Jack Cummins, Sayantan Kumar, Ketan Tamirisa, Jeremy C. Weiss Published: 2026-09-11 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.13475v1 Existing pipelines for clinical timeline extraction from case reports are evaluated using an expert reference and are limited by imperfect reference annotations and imprecise event alignment. We developed GAVEL, an LLM judge protocol that compares two timelines with the case report and returns a discrepancy type, verdict, and report passage for each difference. We evaluated the event matcher, reviewed 2,738 findings from GPT5.6sol and DeepSeek V3.2, ranked six LLM extractors and two human annotators, and tested GAVEL guided merging. True match rates were 60% immediately below and 48% immediately above the 0.10 cutoff. Manual review confirmed 89.4% and 88.6% of findings. Across 126 reports, merged timelines were preferred in 77.0% of comparisons (95% CI, 69.8 to 84.1%) and reduced discrepancies attributed to the evaluated timeline from 7.63 to 0.85 per report. GAVEL supports report-based comparison and revision without treating either timeline as ground truth. --- ### OrchSLM: Probing the Dynamics of Small Language Model Orchestration Authors: Chengxi Zhang, Yu Yao Published: 2026-09-11 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.13470v1 Although large language models (LLMs) have demonstrated remarkable capabilities, their reliance on cloud-scale infrastructure poses fundamental challenges for deployment in agentic pipelines, including latency, privacy, connectivity, and substantial computational cost. Small language models (SLMs) offer a compelling alternative: recent studies suggest that many repetitive and narrowly scoped subtasks in agentic workloads may be better served by specialized SLMs than by monolithic LLMs. However, the limited capacity and context windows of SLMs can constrain long-horizon reasoning and interaction-heavy orchestration strategies such as iterative verification and debate. This motivates a complementary, non-interactive paradigm in which heterogeneous SLMs independently generate candidate solutions and a router orchestrates their cached samples without further model interaction. To further understand the mechanisms of such orchestration, we introduce OrchSLM, a routing framework that unifies existing non-interactive orchestration methods and exposes their underlying design choices as controllable parameters. Using OrchSLM as a systematic probe, we reveal how orchestration behavior emerges from diverse knobs, including the task structure, model-pool composition, and multi-agent consensus. --- ### Inverse Learning of the Altruism and Cost Level in Mixed-Individual Mean Field Games Authors: Haoyang Cao, Gökçe Dayanıklı, Xiaofei Shi Published: 2026-09-11 Categories: math.OC, cs.LG arXiv: https://arxiv.org/abs/2609.13469v1 Understanding how humans respond to incentives, both at the individual and collective levels, is crucial to the design of effective policies. Within the continuous-time stochastic framework for large interacting populations, mean field games (MFGs) model populations of non-cooperative agents, whereas mean field control (MFC) describes the fully cooperative benchmark, interpreted in our setting as fully altruistic behavior. Mixed-individual MFGs interpolate between these two extremes through a parameter governing the degree of altruism. A central challenge for regulators and policymakers, however, is that intrinsic altruism levels and other private structural parameters, such as individual labor costs, are typically unobservable. To address this challenge, we develop an inverse learning framework for mixed-individual MFGs. Our approach enables the recovery of (latent) altruism and labor cost levels from noisy observations, with experiments demonstrating the feasibility and accuracy of our method. These findings underscore the promise of inverse MFG methodologies for uncovering latent preference structures in large populations, with important implications for incentive design, empirical behavioral modeling, and data-driven policy analysis. --- ### Symmetry- and Property-Aware Crystal Generation with Reinforcement Learning for Inverse Materials Design Authors: Ting-Wei Hsu, Arun Bansil, Qimin Yan Published: 2026-09-11 Categories: cond-mat.mtrl-sci, cs.AI, physics.comp-ph, physics.optics arXiv: https://arxiv.org/abs/2609.13468v1 The inverse design of crystalline materials ultimately seeks structures with desired physical properties. However, for many functional responses, a favorable numerical value is meaningful only when supported by the symmetry of the underlying crystal. Without the appropriate crystallographic constraints, an apparent response may be ill defined, accidental, or not symmetry protected. Here we introduce SPARC, a symmetry- and property-aware reinforcement learning framework that optimizes physical objectives while preserving the structural conditions required for their realization. We demonstrate SPARC on two complementary tasks. The first targets strong uniaxial dielectric anisotropy, a tensorial response that is well defined only within appropriate crystal classes. The second maximizes the spectroscopic limited maximum efficiency, a scalar device-level objective without a prescribed symmetry class, allowing the framework to identify favorable crystallographic motifs. These results show that symmetry is not merely an additional design constraint, but a physical foundation for generating candidates with meaningful, robust, and realizable functional properties. --- ### Governing at Machine Speed: An Adaptive Intelligence Architecture for Real-Time AI Policy Enforcement Authors: Sandeep Bokkasam, B. Durgalakshmi Published: 2026-09-11 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.13466v1 Enterprise AI adoption has reached 78% of organizations globally, yet the infrastructure to govern that adoption has not kept pace. This paper identifies and characterizes the attestation deficit, a structural condition in which organizations maintain governance policies but cannot produce auditable, tamper-evident evidence of enforcement within regulatory timelines. Drawing on empirical data from the Stanford 2026 AI Index Report (362 documented incidents), the IBM/Ponemon 2026 Cost of a Data Breach study (USD 4.99M average cost, 92% lacking access controls), and the EY/AIUC-1 Consortium survey (38% end-to-end monitoring, 17% agent-to-agent coverage), this paper demonstrates that the governance failure is organizational and architectural rather than technical. To address this deficit, we propose AGIL (Adaptive Governance Intelligence Layer), a conceptual five-layer architecture designed to use machine learning for real-time AI governance enforcement. The proposed layers include: (1) Autonomous Discovery for shadow AI detection via behavioral fingerprinting, (2) Behavioral Risk Classification unifying security, hallucination, privacy, and accountability scoring, (3) a Policy Enforcement Gateway for inline permit/deny/modify decisions at sub-100ms latency, (4) a Continuous Attestation Engine generating tamper-evident audit trails as a byproduct of enforcement, and (5) Adaptive Policy Intelligence for ML-driven policy evolution across jurisdictions. AGIL is presented as a theoretical framework and architectural proposal; empirical validation through controlled deployment remains a direction for future work. --- ### Root-Cause Attribution Is a Search Problem: Continual Search for Long-Horizon Agent Failures Authors: Harsh Raj, David Lee, Anas Mahmoud, Renxiong Wang, Razvan-Gabriel Dumitru, Chenguang Wang, Tong Zhao, Yunzhong He, Darvin Yi, Vipul Gupta Published: 2026-09-11 Categories: cs.AI, cs.HC, cs.LG, cs.SE arXiv: https://arxiv.org/abs/2609.13463v2 The increasing deployment of AI agents in long-horizon tasks yields massive execution logs. Diagnosing failures within these records is crucial for reliability, as it transforms outcome-level signals into actionable interventions. The sheer scale of the data renders human review impractical, driving the need for automated root-cause attribution (RCA). However, automated RCA methods using LLMs suffer from low diagnostic accuracy, especially as execution traces grow larger. They struggle because relevant information is often sparse, distributed across distant actions, and disconnected from the visible failure, reducing root-cause attribution to a massive search problem. Existing RCA methods typically rely on one-shot LLM judgments to diagnose failures from execution traces. While effective for shorter trajectories, these judges tend to settle on a plausible diagnosis early, leaving critical evidence in longer traces unexamined. We introduce Continual Search, an iterative framework that nudges the judge, over successive turns, to keep searching for unresolved diagnostic evidence. We evaluate Continual Search across four existing RCA benchmarks. Recognizing the lack of massive execution traces in current benchmarks, we introduce MegaRCA-Mix to evaluate RCA at scale. MegaRCA-Mix provides a challenging testbed of 50 human-annotated failure trials spanning long-horizon, execution-heavy tasks. Across multiple benchmark suites and model families, Continual Search consistently improves attribution performance. On MegaRCA-Mix, for example, it improves Opus-4.8's F1 score by 29\%, from $0.471$ to $0.608$. More interestingly, within the same model family, lower-tier models can even surpass their higher-tier counterparts, demonstrating that effective search supersedes raw model scale. --- ### STAGE: Diagnosing Semantic Transfer at Grounded Execution in Embodied Agents Authors: Baosheng Jin, Yushen Liang, Hua Shen Published: 2026-09-11 Categories: cs.RO, cs.CL arXiv: https://arxiv.org/abs/2609.13458v1 Embodied language grounding requires more than identifying the referent of an instruction: recovered semantics must also control the action an agent exposes. We study this missing link as a semantic-action gap, where instruction semantics are recoverable but weakly expressed in native continuous actions. We introduce SAT-Bench, a fixed-observation counterfactual benchmark that holds the visual scene and agent state fixed while changing only instruction semantics. On LIBERO target-name and pixel-grounded relation swaps, target recovery reaches 100.0% and 95.8%, whereas OpenVLA action sensitivity remains only 6.8% and 7.7%. The gap persists across 1,000 additional compositional and temporal/procedural counterfactuals, with overall action sensitivity of 6.1%. Hidden-state, threshold-free, cross-policy, and rollout diagnostics further support this semantic-action transfer failure. We introduce VISA, a lightweight execution-time interface that converts recovered semantics into ALLOW, DEFER, target-consistency, and verified-selection decisions. VISA reduces invalid-instruction blind execution from 92.7% to 2.8% while preserving 94.0% of normal commands, and verified selection further improves target-consistent action exposure without updating the underlying policy. Overall, embodied language evaluation should measure semantic-action transfer, not semantic parsing alone. --- ### TimeThink: Eliciting Compositional Reasoning in Timeseries Large Language Models Authors: Sudarshan Regmi, Arvind Pillai, Yu Yvonne Wu, Yuliang Chen, Bibek Panthi, Tess Z. Griffin, Michael V. Heinz, Lisa Marsch, Nicholas C. Jacobson, Andrew Campbell Published: 2026-09-11 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.13457v1 Timeseries multimodal large language models (TS-MLLMs) have recently begun leveraging the reasoning capabilities of large language models (LLMs) for question-answering tasks. However, these models often fail to capture dynamic temporal patterns, providing only implicit reasoning that lacks the underlying explanations critical for high-stakes applications like healthcare. While reinforcement learning (RL)-based timeseries language models aim to address this, they often fall short because they are trained on narrow, in-distribution data and struggle with out-of-distribution compositional questions. To address these challenges, we present TimeThink, a synthetic framework for eliciting compositional timeseries reasoning. Core timeseries primitives (e.g., trend, seasonality) are domain-independent and can be deterministically generated. Guided by this premise, TimeThink first designs a synthetic data generator that produces atomic and composite question-answer pairs, providing objective ground truth with reasoning traces. Building on this framework, TimeThink employs a reinforcement learning with verifiable rewards (RLVR) training strategy that encourages explicit reasoning. Unlike template-reliant methods, this approach enables the model to learn the underlying logic of composition rather than simply imitating traces. Extensive experiments show that TimeThink, trained only on synthetic data, significantly outperforms strong baselines on both synthetic and real-world benchmarks. --- ### Hindsight Bias in Clinical Temporal Reasoning: How Future Data Exposure Affects Large Language Model Judgment Authors: Misaki Matsuura, Sayantan Kumar, Ojas Kadam, Jeremy C. Weiss Published: 2026-09-11 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2609.13454v1 Clinical decisions are prospective, but clinical language models are often evaluated on retrospective records that reveal the final diagnosis, treatment response, and outcome. Such evaluations may reward the use of future information rather than reasoning under the uncertainty present at the decision point. We introduce a paired benchmark for measuring outcome-conditioned shifts consistent with hindsight bias in clinical temporal reasoning. It contains 171 case reports from the PubMed Central Open Access Subset---40 sepsis and 131 GLP-1/diabetes cases---represented as both textual narratives and human-annotated and LLM-generated textual time series (TTS). For each case, questions are tied to a clinically meaningful cutoff and paired with a prospective reference answer and an outcome-consistent \emph{hindsight trap}. Models answer each question using either a TTS truncated at the cutoff or the complete timeline; additional conditions vary the narrative source (original or synthetic) and TTS annotation source (human or LLM). We evaluate accuracy (Acc), hindsight trap rate (HTR), answer instability rate (AIR), and hindsight bias rate (HBR), each of which captures different signals of hindsight bias. Across GPT 5.6 Sol, Gemma 4, GLM 5.2, and Opus 5, full timeline exposure produces consistent hindsight-sensitive shifts, while temporal masking reduces bias without lowering accuracy. --- ### A Machine Learning API for Earth Observation Data Cubes Based on openEO Authors: Brian Pondi, Jonas Hurst, Rolf Simoes, Jonas Starke, Marius Appel, Edzer Pebesma Published: 2026-09-11 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.13453v1 Earth Observation (EO) data are increasingly organized as spatio-temporal data cubes, while machine learning (ML) methods operate on tabular feature matrices or structured tensor inputs. This mismatch forces platform-specific transformations that are difficult to reproduce or transfer across cloud infrastructures. The openEO specification provides a unified interface for EO data access and processing across heterogeneous backends, but lacks a standardized approach for ML integration. We propose a process-level ML specification for openEO structured into three stages: model initialization, model actions (training, tuning, inference, validation), and model management. It supports classical algorithms such as Random Forest and SVM, as well as deep learning architectures for time-series and spatial patch-based modeling, including TempCNN, Temporal Attention Encoders, and foundation model inference. Three prototype implementations in R and Python demonstrate feasibility across diverse technology stacks. A crop type mapping use case demonstrates cross-backend interoperability by submitting an identical process graph to independent R and Python backends and comparing predictions and evaluation metrics. Two further use cases demonstrate deep learning on time series and foundation model inference, each executed on a dedicated backend. The prototypes reveal, however, that full cross-backend portability requires deeper harmonization of serialization formats and execution semantics than the process level alone can enforce; backend library versions and preprocessing conventions outside the specification's boundary also affect reproducibility. Addressing both through explicit backend conformance profiles represents the most important near-term direction. The specification advances the reproducibility, portability, and accessibility of ML workflows on EO data cubes across cloud platforms. --- ### Causal Analysis and Mitigation of Spurious Onsets in Full-Duplex Speech LLMs Authors: Kento Nishi Published: 2026-09-11 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.13445v2 Speech-to-speech LLMs like Moshi, and its derivative PersonaPlex, can listen and speak concurrently through full-duplex generation. However, they can begin speaking inappropriately during prolonged user silence: under digital-zero input, Moshi and PersonaPlex initiate speech in 30% and 27.5% of five-minute continuations, respectively. What causes this spurious speech? We investigate two hypotheses: either repeated sampling selects speech despite persistently low onset probabilities, or self-conditioning on nonspeech outputs causes an abrupt spike in onset probability. We find that, at every observed onset, speech probability spikes by over nine orders of magnitude in one 80-ms frame, supporting the latter hypothesis. Then, to suppress these onsets without blocking genuine responses, we ask a causal counterfactual question: is the model responding to user speech, or would its next-token distribution remain similar if the preceding user input were muted? Accordingly, we suppress onsets whose distributions change little under this intervention. Under realistic microphone noise, our method suppresses spurious onsets, while preserving genuine responses: one-sided 95% lower confidence bounds are 98.68% and 98.82% for Moshi, and 96.90% and 99.25% for PersonaPlex. Our inference-time method runs in real-time without retraining, with 95th-percentile decision time below 61 ms, within the 80-ms frame budget. Our code is available at https://github.com/KentoNishi/icassp27-spurious-onsets. --- ### Learning to Solve Hard Problems in RL for LLMs by Never Giving Up Authors: Michael Noukhovitch, Hamish Ivison, Nathan Lambert, Aaron Courville Published: 2026-09-11 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.13443v1 We demonstrate that training LLMs with RL does not improve performance equally across a dataset. RL shows large improvements on easy problems that an LLM is already good at solving, but small improvements on hard problems. We call this the Matthew Effect in RL for LLMs, after the phenomenon of cumulative advantage from economics and network science summarized as "the rich get richer". The naive explanation is that hard problems require more compute to find a solution. We argue that modern RL methods are exacerbating the issue by wasting too much compute on easy problems and instead should dynamically reallocate how they use compute. We introduce Never Give Up (NGU), a simple adaptive sampling method that keeps generating samples for a problem until one is correct. By leveraging asynchronous RL, this naturally uses fewer samples to filter out easy problems and allocates more compute to solving harder problems. We investigate the design choices that affect NGU, such as off-policy robustness, and develop a set of best practices. On the math benchmark Deepscaler, NGU improves performance per compute, especially on harder problems. On a recent coding task, Manufactoria, standard GRPO with a per-test reward fails to fully solve problems that have a range of easy and difficult tests. NGU iteratively improves, solving harder and harder tests, until it learns to fully solve coding problems. --- ### Efficient Online Inverse Optimization with $O(d)$ Regret Authors: Yang Cai, Anupam Gupta, Vineet Gupta, Guru Guruganesh, Yanchen Jiang, Christopher Liaw, Aranyak Mehta, Renato Paes Leme, Grigoris Velegkas, Di Wang Published: 2026-09-11 Categories: cs.LG, cs.DS arXiv: https://arxiv.org/abs/2609.13440v1 We give a deterministic algorithm for online inverse linear optimization with regret $O(d)$, uniform in the horizon and $O(d^{2})$ time per round. A bound of this order was obtained recently by Dewasurendra, settling a question of Gollapudi et al.\ and of Oki and Sakaue, but by an improper rule that enumerates covers at every scale and costs $T^{Θ(d)}$ a round; ours is the first efficient such bound and the first proper one. We build on the variable-metric framework of Sakaue et al., adding a self-normalized rank-one update, and we replace the $\log\det$ potential by the trace power $\tr(H^{-1/2})$, which is bounded outright and removes the $\ln T$. The bound also holds against an expert that does not optimize, and we give corruption-robust and rank-adaptive variants, and an application to convex minimization. --- ### Certifiably Interpretable Training of ReLU-MLPs for Boolean Tasks with Guaranteed Truth-Table Generalization Authors: Hrad Ghoukasian, Anastasis Kratsios Published: 2026-09-11 Categories: cs.LG, cs.AI, stat.ML arXiv: https://arxiv.org/abs/2609.13439v1 As compute scales, models evolve, and training algorithms advance, our ability to explain the increasingly powerful AI systems they enable is eroding. To help safeguard interpretability, we introduce a specialized training algorithm (MACCHIATO) that jointly constructs (i) an explicitly structured $\operatorname{ReLU}$-MLP from partial truth-table observations and (ii) an explicit Boolean circuit over signed literals with $\{\operatorname{AND},\operatorname{OR},\operatorname{XOR}\}$ gates certifying what its subnetworks compute and how they compose. Intuitively, we iteratively project the residuals of a Boolean function onto low-dimensional $\{\operatorname{AND},\operatorname{OR},\operatorname{XOR}\}$-circuit classes and exactly compile the resulting circuit into a $\operatorname{ReLU}$-MLP; we combine $\operatorname{ReLU}$-MLP circuit compilation, ESPRESSO logic minimization, and influence-based variable selection. Roughly speaking, our interpretability certificate is complemented by a statistical guarantee: under the theorem's influence-recovery conditions, if each of the $m$ stage-wise residuals depends on at most $\log_2(B)$ bits, a sample-splitting variant of our algorithm trained on $T$ observations returns a six-layer $\operatorname{ReLU}$-MLP (counting the input layer) of width $\mathcal{O}(mB)$ with truth-table error $\mathcal{O}\bigl(\sqrt{m(B+\log(m/δ))/T}\bigr)$. On synthetic random-junta tasks, our networks outperform depth- and hidden-width-matched Adam-trained MLPs in several data-sparse or projection-aligned regimes, while the trained ReLU-MLPs are stronger in others. Moreover, in our explicit PyEDA truth-table implementation, the iterative procedure completes in regimes where flat ambient-dimensional ESPRESSO exceeds the three-hour computational budget. --- ### LabAgent: Customize Any Research Hubs for Scientific Discoveries Using AI Agents Authors: Lei Liu, Yikun Zhang, Jialin Chen, Wanjia Zhao, Rex Ying, Wengong Jin, Hua Xu, James Zou, Tianyu Liu, Hongyu Zhao Published: 2026-09-11 Categories: cs.AI, q-bio.BM arXiv: https://arxiv.org/abs/2609.13437v1 Scientific research is a continuous process that emphasizes inheritance. Methods developed by predecessors are often expanded upon by new researchers to explore more novel and in-depth scientific questions. However, the change of lab staff, such as student graduation, leads to a lack of personnel capable of replicating methods. Methods that have been developed with significant effort and resources cannot be continued. To address these limitations, we propose LabAgent, a reproduce and discovery harness tailored for a lab's continuous work. LabAgent employs two mechanisms to guarantee that all skills can be executed and verified and to record the corrective methods and experiences, allowing for direct correction or avoidance of similar errors. We applied LabAgent to drug property prediction, biomedical problem analysis, protein variant effect prediction, and statistical genetics in life science domains. LabAgent ranks first over commercial generalist agents in every domain, and demonstrates accurate reproduction of a published figure. Overall, these results demonstrate that LabAgent can effectively integrate and reasonably expand laboratory knowledge. --- ### Toward Self-Adaptive Physical AI: Can LLM Agents Manage Long-Horizon Physical Tasks? Authors: Varun Kaushik, Yayun Tan, Xiaofan Yu Published: 2026-09-11 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.13436v1 Large Language Model (LLM) agents offer a promising path toward autonomously managing long-term physical tasks without human intervention. However, physical tasks require agents to continuously observe the environment, make consequential actions, and remain effective as the environment changes. Existing approaches either require substantial data and retraining, or primarily focus on agents operating in the virtual world. In this work, we explore the feasibility of building a self-adaptive physical AI agent that manages long-term physical tasks in a zero-shot manner and adapts to environmental changes without human intervention. We design a multi-agent framework that integrates planning, tool calling, observation, and verification, and evaluate it on agricultural tasks against reinforcement learning (RL) agents under different weather patterns. Our results show that zero-shot LLM agents can achieve comparable management outcomes to RL agents under the same weather pattern and adapt more effectively than RL when evaluated under a shifted environment, highlighting a promising path toward self-adaptive physical AI agents. --- ### Chance-Constrained Belief-Space Maneuver Planning for Autonomous Collision Avoidance Under Uncertainty Authors: Grace Ra Kim, Duncan Eddy, Mykel J. Kochenderfer Published: 2026-09-11 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2609.13428v1 Increasing conjunction frequency in low Earth orbit places growing pressure on spacecraft operators to determine not only whether an encounter requires mitigation, but whether sufficient information is available to commit to a maneuver. This work formulates this information-action tradeoff as a belief-space planning problem for conjunctions between a maneuverable spacecraft and an unmaneuverable secondary object. The planner represents the uncertain orbital states as Gaussian beliefs and uses a chance-constrained belief-space Monte Carlo tree search framework to reason over possible future tracking updates before time of closest approach (TCA). A terminal chance constraint limits the probability of reaching TCA above a prescribed collision-risk threshold, allowing the planner to wait for informative tracking while intervening when deferral becomes too risky. We evaluate the approach on eight historical conjunctions from NASA's Conjunction Assessment Risk Analysis dataset. By varying the secondary-object measurement quality and tracking cadence, we generate a total of 96 distinct evaluation scenarios. Across the evaluated conditions, the planner reaches TCA without maneuvering in approximately 40% of episodes while maintaining no terminal collision-risk violations. In contrast, fixed-time rule-based maneuver policies resolve more encounters without maneuvering when intervention is deferred closer to TCA, but at the expense of increasing terminal risk violations. The fraction of episodes reaching TCA without maneuvering depends strongly on tracking quality and measurement cadence, ranging from 76% under accurate, frequent measurements to approximately 18%-20% under the poorest tracking conditions. These results show that tracking quality and frequency are not only inputs to collision-risk estimation: they can determine when intervention becomes necessary. --- ### ReCAST: Reward Credit Assignment across Timesteps for Online Diffusion Reinforcement Authors: Yihang Chen, Yuanhao Ban, Kuei-Chun Kao, Cho-Jui Hsieh Published: 2026-09-11 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.13425v1 Training diffusion models with multiple rewards requires distinguishing user preference from reward informativeness. User preference determines how much each reward should contribute to the overall objective; reward informativeness determines when its feedback is useful during denoising. Some rewards can meaningfully evaluate a sample as soon as global structure emerges, but others become informative only when the sample is nearly clean. To address both questions jointly, we propose ReCAST (Reward Credit ASsignment across T}imesteps), the first method, to our knowledge, for per-reward, timestep-dependent credit assignment in diffusion reward fine-tuning. ReCAST separates user preferences from temporal allocation through a reward-by-timestep weight matrix $W$, whose row sums match the user-specified reward budgets $λ$, while its column sums are equal, assigning the same total weight to each denoising step. Under these marginal constraints, ReCAST allocates weight according to each reward's informativeness, quantified by its Rényi discriminability gain at each step. These gains telescope to the total discriminability between the reward-induced positive policy and the current policy, providing a basis for temporal credit assignment. We evaluate ReCAST by training SD3.5-Medium under two distinct four-reward settings, each across five reward budgets $λ$. ReCAST improves the training rewards in one setting and matches them in the other, improves every held-out judge in both, and is preferred by an independent LLM-as-a-Judge. Together, these results show that ReCAST yields improvements that generalize beyond the training rewards and support its core principle: assigning each reward greater weight at the denoising timesteps where its feedback is most informative. --- ### Vibe Patenting: Evaluating LLM Judges for Professional Patent-Drafting Agents Authors: Toshiaki Koike-Akino, Vladislav Blaykhman, Ye Wang, Jing Liu, Gene V. Vinokur Published: 2026-09-11 Categories: cs.AI, cs.LG, cs.MA arXiv: https://arxiv.org/abs/2609.13422v2 LLM judges are increasingly used to evaluate and improve AI-generated outputs, yet their reliability for complex professional work remains unclear. We study this problem through Vibe Patenting, an end-to-end patent-drafting testbed for AI-agent evaluation. A separately-invoked LLM judge evaluates generated patent drafts and provides structured feedback for iterative revision. Across multiple inventions and drafting-agent configurations, judge-guided revision consistently improves judge-assessed quality, while unguided revision tends to saturate. Notably, iterative judge feedback enables a low-reasoning agent to approach the performance of a substantially more expensive high-reasoning agent. Stronger models and increased reasoning generally improve judge-assessed drafting quality, while domain-specific agentic workflows provide further gains. We validate the judge against independent evaluation by a professional patent attorney and find meaningful but strongly metric-dependent agreement and systematic calibration differences. These results highlight both the utility and limitations of LLM judges as evaluators and optimization signals for complex professional workflows. --- ### Specification Oracles Authors: Atticus Cull, Justin McCarthy Published: 2026-09-11 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.13415v1 Specifications face a basic tradeoff: leave details out, and important questions go unanswered; record every detail separately, and the specification becomes large and prolix. We investigate whether a language model can serve as a compact, living specification oracle by learning facts about a target and answering questions about it directly. We compare two ways of storing the learned facts: external text notes and changes to the model's weights. Across four families of 596-fact worlds and two Qwen2.5 model sizes, weight-only oracles benefited substantially more from structure: with the 7B model, their accuracy integrated across storage capacities was 18.5 percentage points higher on structured than unstructured worlds, compared with 1.1 points for note-sheet oracles. This advantage came at a substantial storage cost, with the smallest adapter requiring approximately 175 KiB compared with a maximum note budget of 16 KiB. Adapted weights therefore exploited latent structure more successfully, while external notes required substantially less object-specific storage. --- ### CVSS-X: A Multilingual Speech-to-Speech Translation Corpus for 28 Languages Authors: Lucas Rafael Stefanel Gris, Alef Iury Siqueira Ferreira, Frederico Santos de Oliveira, Augusto Seben da Rosa, Alexandre Costa Ferro Filho, Arlindo Rodrigues Galvão Filho, Anderson da Silva Soares Published: 2026-09-11 Categories: cs.CL, cs.SD, eess.AS arXiv: https://arxiv.org/abs/2609.13413v2 We introduce CVSS-X, a large-scale synthetic speech-to-speech translation corpus that extends CVSS by reversing the translation direction. While CVSS translates from 21 languages into English, CVSS-X enables translation from English into 28 target languages spanning 12 language families. The corpus comprises approximately 240,000 parallel speech pairs per language, totaling over 16,000 hours, eight times larger than CVSS. We provide two variants: CVSS-X-C with two canonical voices per language, and CVSS-X-T with cross-lingual voice cloning, both fully generated. Evaluation shows comparable translation quality to CVSS with consistent performance across typologically diverse languages. Combined with CVSS, this enables research on bidirectional and multilingual speech-to-speech translation. The code is available at https://github.com/ErmisAI/XVSS-X and the dataset under CC-BY-NC 4.0 license at https://huggingface.co/datasets/lgris/XVSS-X. --- ### Translation-Invariant Tile-Based Phase Unwrapping with Residual-Weighted Multipath Averaging Authors: Antoine Moevus, Max Mignotte Published: 2026-09-11 Categories: eess.IV, cs.CV arXiv: https://arxiv.org/abs/2609.13409v2 Phase unwrapping is a key step in interferometric and coherent imaging, where the physical quantity of interest is carried by a phase that the instrument delivers only modulo 2*pi. The difficulty in two dimensions is to separate the jumps caused by wrapping from those produced by noise, by true discontinuities, by under-sampling or by decorrelation. Spatial-domain and frequency-domain methods have both been studied extensively, each with advantages the other lacks; hybrid schemes combining the two remain scarce. We propose a semi-global tile-based strategy in which every tile is unwrapped in the frequency domain, through the Discrete Cosine Transform (DCT) and the least squares (LS) formalism of Ghiglia et al., the tiles being merged spatially. Unwrapping and noise filtering are performed jointly, which regularizes an otherwise ill-posed inverse problem. An error stays confined to the tile in which it arose; the tiling artifacts are removed by averaging over every shift of the grid and over the symmetries of the square; and that average is weighted by the Poisson residual each pass leaves behind, so that a pass whose tile boundaries fell on a discontinuity does not impose its seam on the result. Experiments on synthetic and real data, against four reference algorithms from four distinct families and under six complementary metrics, show that the proposed method matches or improves on the state of the art. One of these metrics, a corrected cyclic re-wrap residual introduced here, needs no ground truth and therefore remains available on real acquisitions. --- ### Generalized Agent Iteration: One Formal Framework for Iterative Policy Improvement and Recursive Self-Improvement Authors: Hongyao Tang, Yi Ma, Pengyi Li, Yifu Yuan Published: 2026-09-11 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.13406v1 When we speak of recursive self-improvement (RSI), are we speaking of a phenomenon, a mechanism, or a prospect? Towards autonomous and evolving intelligence, RSI is being claimed at many scales, while no single framework that formally describes these emerging instances exists. Its counterpart in the classical realm, iterative policy improvement, is characterized by generalized policy iteration (GPI), a framework of broad applicability with well-understood theoretical properties, but only where the update principle and the evaluation base lie outside the agent. In this paper, we propose Generalized Agent Iteration (GAI), a formal framework that describes iterative policy improvement and RSI as two cases of a single learning paradigm. GAI defines the agent as a configuration of modifiable components within a system and models the learning process as a cycle of agent evaluation and agent improvement. Two pivotal dials then distinguish the instances: whether the improving mechanism is part of the agent and whether the standard it is measured against is grounded outside it. The former dial delineates the boundary between GPI and RSI, and the latter determines a system's polarity as anchored, goal drift, or fully self-referential. Moreover, we use these coordinates to place existing systems on the same two axes and make the defects of recursive self-improvement statable one condition at a time. We see this paper as a first step toward exploring a formal characterization of RSI that rests on the classical account, makes existing systems comparable, and provides a principled basis for analyzing and designing new ones. --- ### ConeGaussian: Anti-Aliased Gaussian Ray-Tracing for Generic Central Cameras Authors: Deheng Zhang, Letian Shi, Runyi Yang, Zhendong Li, Lei Sun, Kanzhi Wu, Ajad Chhatkuli, Danda Pani Paudel, Luc Van Gool Published: 2026-09-11 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.13397v1 In rendering, a camera is a sampling operator that maps each finite pixel to a bundle of rays. Different camera models change the geometry of this bundle, thus making a unified and faithful rendering formulation challenging. Consequently, Gaussian ray tracing supports generic cameras (with optical center) through their inverse ray mappings, yet typically reduces every pixel to a single center ray. This ignores the camera-dependent pixel footprint, causing aliasing under minification, while unconstrained Gaussians expose unsupported frequencies under magnification. We present ConeGaussian, a camera-model-agnostic anti-aliasing framework for Gaussian ray-based rendering. Instead of defining the pixel filter on a camera-specific image plane, ConeGaussian constructs an anisotropic footprint directly from neighboring rays produced by the camera's native inverse mapping. We derive a closed-form response under a locally linear, depth-local, moment-matched approximation of the finite pixel footprint, while the same geometry defines a per-Gaussian training-frequency floor. Notably, by construction, our filtering principle can be used unmodified across calibrated central camera models and multiple Gaussian ray-rendering backbones. Additionally, unlike in mip-splatting, our scene-space frequency floor and filtering enable trivial composition at render time, allowing us to remove excess blurring. On pinhole and strongly distorted fisheye captures, ConeGaussian consistently improves two distinct ray-based backbones, by up to 4.3 dB at 1/8 resolution, and reduces fisheye LPIPS by 30% where perspective screen-plane footprint formulations are not directly applicable. --- ### Converge Then Diversify: Decoupling Convergence and Diversity in Multi-Objective Bayesian Optimisation Authors: Chao Jiang, Yueling Huang, Miqing Li Published: 2026-09-11 Categories: cs.AI, stat.ML arXiv: https://arxiv.org/abs/2609.13396v1 Multi-objective Bayesian optimisation (MOBO) is a sample-efficient approach for optimising expensive black-box functions with multiple objectives. In MOBO, the goal is to adequately approximate the Pareto front; that is, to obtain a high-quality solution set with 1) good convergence (closeness to the Pareto front) and 2) good diversity (spread across the Pareto front). Existing MOBO methods typically aim to accomplish these two tasks simultaneously, i.e., driving the search towards the Pareto front while maintaining a diverse set of nondominated solutions, such that the solutions, ideally, can gradually approach the entire front. When sufficient search budgets are available, this approach is effective. However, considering both convergence and diversity throughout the search is not easy and requires careful design. Under very tight budgets, there may not be enough solutions generated to be able to simultaneously approach the entire Pareto front. To address this issue, this paper proposes a \textit{converge-then-diversify} (CTD) approach that decouples convergence and diversity into two stages. In the first stage, CTD focuses on convergence, aiming to quickly drive the search toward a single point on the Pareto front. In the second stage, CTD focuses on diversity, aiming to spread solutions across the front. We present two simple instantiations of CTD by using widely adopted acquisition functions in the area. Experimental results show that, across all 446 pairwise comparisons, CTD statistically outperforms state-of-the-art methods in 72.9\% of the cases, performs equivalently in 21.1\%, and is statistically worse in only 6.1\%, with the advantage being particularly evident in settings with very tight evaluation budgets or in high-dimensional problems. ---