# Research Papers — 2026-09 · bundle page 84 of 266 > Full-text markdown of papers #4151–4200 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/84 [← Previous bundle](https://aipod.dev/papers/archive/2026-09/page/83.md) [Next bundle →](https://aipod.dev/papers/archive/2026-09/page/85.md) ### Cross-Modal Knowledge Distillation for Acoustic Pedestrian Detection Authors: Yonghyun Kim, Chaeyeon Han, Sancho Gatungay, Subhrajit Guhathakurta, Alexander Lerch Published: 2026-09-24 Categories: eess.AS, cs.LG, cs.SD arXiv: https://arxiv.org/abs/2609.31785v1 Audio-only pedestrian detection is attractive for urban sensing but limited by weak acoustic cues. An appealing strategy is cross-modal knowledge distillation, in which a video teacher supervises the audio student during training so that the deployed model runs on audio alone. Under this task's severe class imbalance and wide video-audio modality gap, however, what such distillation contributes is unclear. We introduce Trust-Filtered Distillation (TFD), which selectively suppresses teacher supervision on pedestrian samples, and interpret its logit formulation as conditional label smoothing under a shared temperature. Across ten distillation configurations under five-fold cross-session validation on ASPED, most methods yield modest gains in macro accuracy accompanied by small changes in PR-AUC. The main effect is higher no-pedestrian accuracy at the cost of lower pedestrian recall. Adding TFD to logit distillation strengthens this trade-off without improving mean PR-AUC. These findings clarify the benefits and limitations of selective cross-modal supervision by distinguishing operating-point shifts from discrimination gains in imbalanced acoustic detection. --- ### Witeness Overlap: Directional Provenance Inside Open-Weight Model Families Authors: Siyuan Li, Haoxuan Zeng, Xin Luo, Fernando Jia, Florence Li, Zhengyang Geng, Zico Kolter, Tai Sing Lee, Tianqin Li Published: 2026-09-24 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.31784v1 Open-weight models are often released, fine-tuned, aligned, merged, and re-released, making provenance audits ask not only whether checkpoints are related, but also which checkpoint came first. Many existing model-provenance methods are designed for a base-known audit setting: given a victim or source model, they test whether a suspect model is related to it. Although these audits are framed as source-to-suspect tests, their underlying evidence is often symmetric, relying on representation similarity, weight similarity, behavioral fingerprints, or correlation statistics. Symmetric pairwise comparisons can detect relatedness, but they cannot by themselves orient relationship between checkpoints A and B. We therefore introduce a local geometric comparison: instead of comparing two checkpoints directly, we add a third same-family checkpoint as a witness and compare the geometry around each candidate endpoint. Direction is inferred by asking which candidate behaves more like a branching parent. Motivated by this idea, and by the empirically observed asymmetry between parent-anchored and child-anchored witness-overlap distributions, we propose Witness Overlap, a prompt-free, training-free white-box test for directional provenance. On 176 LLM checkpoints from 16 families, our one-witness test orients 95.3\% of parent-child decisions using Frobenius cosine. We further evaluate root identification, sibling discrimination, generalizations to VLM and diffusion families, and chain-structured ordering. The signal is robust to weight noise and sparse pruning, with a proposed SVD weight reduction variant showing greater robustness than Frobenius cosine. --- ### Panoptic Scene Program Diffusion Transformer Authors: Chika Maduabuchi Published: 2026-09-24 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2609.31780v1 Modern text-to-image models produce high-fidelity images but still struggle with compositional prompts that require instance identity, attribute ownership, counting, spatial ordering, and role-sensitive relations. We introduce Panoptic Scene Program Diffusion Transformer (PSP-DiT), a diffusion-transformer architecture that treats a panoptic scene program as a first-class latent variable rather than an external control signal or post-hoc parse. PSP-DiT jointly denoises image latents and scene-program latents through coupled transformer streams, while panoptic grounding and cycle-consistency objectives tie object instances, attributes, relations, and counts to visual support in the generated image. Under matched training and inference settings, PSP-DiT improves over a strong flat-text baseline across GenEval 2, SANEval-Simple, PSG-Score, and DetailMaster, with the largest gains on counting, attribute binding, role-sensitive relations, and long structured prompts. The method preserves image quality, adds modest inference overhead, and remains robust to imperfect scene programs. --- ### Beyond MSE: Rician Likelihood Denoising for Self-Supervised Cardiac $T2$ and $T1ρ$ MRI Authors: Nicholas A. Jacobs, Jason Mendes, Ravi Ranjan, Edward DiBella, Shireen Elhabian Published: 2026-09-24 Categories: eess.IV, cs.CV arXiv: https://arxiv.org/abs/2609.31777v1 Magnetic resonance imaging involves an inherent trade-off among spatial resolution, acquisition time, and noise. This trade-off contributes to long scan times and high cost. Deep learning has improved image denoising, but cardiac MRI remains difficult because high-resolution, rapid acquisitions generally lack corresponding low-noise ground truth. Self-supervised denoising offers a potential solution by learning from noisy image pairs or even single noisy acquisitions. However, we show that Noise2Void-style blind-spot denoising, which uses a mean squared error (MSE) loss and assumes zero-mean, independent and identically distributed (i.i.d.) noise, is poorly suited to MR magnitude images. When applied to short-axis $T2$-weighted and $T1ρ$-weighted cardiac MRI with synthetic Rician noise, it produces biased denoised images and biased parametric maps of $T2$ and $T1ρ$. To address this limitation, we formulate self-supervised denoising as maximum likelihood estimation under a known Rician noise model. This yields unbiased denoisers that are competitive with supervised baselines. --- ### Suitable Measures for the Potential Operational Utility of AI NWP Rainfall Forecasts Over Africa Authors: Shruti Nath, Docko Sow, Koomi Toussaint Amoussouvi, Fenwick Cooper, Josiah Kiarie Kimani, John Bagiliko, Florian Pappenberger, Rendani Mbuvha Published: 2026-09-24 Categories: stat.AP, cs.LG, physics.ao-ph arXiv: https://arxiv.org/abs/2609.31775v1 Artificial intelligence (AI)-based weather prediction is approaching the skill of physical numerical weather prediction (NWP) systems at a fraction of the computational cost. This is particularly promising for Africa, where rainfall extremes are intensifying and many forecasting centres lack the infrastructure to run physical models at extended lead times. We present a calibrated comparison of GraphCast, GenCast and the Functional Generative Network (FGN) against the physical NWP model IFS for rainfall prediction across Africa. Deterministic and probabilistic forecasts are postprocessed using Isotonic Distributional Regression and evaluated with the Continuous Ranked Probability Score against IMERG, RFEv2 and CHIRPS across seasons, wet and dry regimes, elevation zones and lead times. All models retain skill beyond climatology across most seasons and at extended lead times. AI models generally outperform IFS in wet regions, whereas IFS performs better in dry, high-elevation areas, where its finer resolution better represents orographic controls on rainfall. Across observational datasets and seasons, AI models achieve a median improvement of approximately 5% over IFS. GraphCast achieves calibrated skill comparable to the ensemble-based FGN, although FGN provides greater significant skill at longer lead times. These results highlight the potential of calibrated AI weather prediction to provide accessible and computationally efficient rainfall forecasts across Africa, while demonstrating the continuing importance of spatial resolution, ensemble design and regional characteristics. --- ### Timed Rule-Based Supervision of an End-to-End Autonomous Parking Policy Authors: Kejia Gao, Liguo Zhou, Lei Yu, Alois Knoll Published: 2026-09-24 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2609.31773v1 We study whether a manually specified runtime supervisor can correct recurring failures of an existing end-to-end parking policy in a fixed CARLA parking lot. The vision-based Transformer architecture is inherited from Yang et al.; our contribution is a timed, rule-based Parametric Safety Shield (PSS) applied to its control outputs. The PSS uses hand-calibrated speed, position, and duration thresholds to intervene in observed failure modes, including boundary exits, delayed braking, and stalled or oscillatory control. In the reported closed-loop evaluation, 16 held-out target slots and six initial poses are each evaluated in four rounds (384 attempts per configuration). Target success increases from 327/384 (85.16%) for the retrained policy to 375/384 (97.66%) with the PSS; mean position and orientation errors among successful attempts are 0.21m and 0.33 degrees. These results show an improvement within this simulator setup. The repeated attempts share one map, vehicle, and sensor configuration, and the PSS uses simulator world coordinates; thus the results do not establish generalization to other lots or real vehicles, or a formal safety guarantee. --- ### PRIME-ANC: Path-Ratio-Informed Modeling for Efficient Neural Filter Synthesis in Active Noise Control Authors: Yaokun Huang, Chunyang Xu, Haowen Hua, Sen Lin, Shichao Hu, Mengyao Zhu Published: 2026-09-24 Categories: eess.AS, cs.AI, cs.SD, eess.SP, eess.SY arXiv: https://arxiv.org/abs/2609.31772v1 Changes in listener acoustics require active noise control (ANC) filters to be redesigned for new acoustic paths. We introduce PRIME-ANC, a shared neural synthesizer that learns a bounded, path-dependent logmagnitude correction to a regularized path-ratio base. Minimum-phase reconstruction and truncation produce finite-impulse-response (FIR) filters; training optimizes their noise-control performance. Across ten random training/test splits per dataset, PRIME-ANC achieves average held-out one-third-octave reductions of 18.81 and 17.76 dB over 50 Hz-5 kHz on a ten-path dataset and a public earphone database, respectively. On original earphone measurements, it improves upon the path-ratio base by 7.89 dB. Ablation studies support the contributions of both the analytic base and the path-dependent correction. Given calibrated paths for a held-out listener condition, PRIME-ANC generates a path-specific FIR without iterative optimization. With three Gauss-Newton updates, it reaches 21.43 dB reduction, approaching direct weighted least-squares design while producing lower amplification and root-mean-square control output. --- ### Robot Manipulation with GPT-6-Astra: Body Knowledge, Experience Reuse, Emergent Skills, and Sim2Real Transfer Authors: Sida He, Lingxi Xie, Yunning Cao, Pengfei Chen, Kaiwen Duan, Jiannan Ge, Xinyue Huo, Jiacheng Shao, Qi Tian Published: 2026-09-24 Categories: cs.RO, cs.AI, cs.CL, cs.CV, cs.LG arXiv: https://arxiv.org/abs/2609.31770v1 General-purpose multimodal agents can write robot-control programs, but repeated exploration and model-mediated action selection can make execution slow. We study how external body knowledge, successful experience, and executable skills improve an XLeRobot controlled by GPT-6-Astra in a simulated and a physical elevator-button task. In 30 fixed-start simulation trials, complete robot geometry and camera information reduce mean completion time by 57.4% relative to a baseline with only the common control interface and no prior experience; images with synchronized action and state records reduce it by 68.6% without additional body assets. In nine paired comparisons (18 trials) at starts displaced by 10-100 cm, experience recorded at the original start reduces mean time by 58-63% relative to no experience, demonstrating generalization to the tested new starting positions. During experience experiments, GPT-6-Astra spontaneously generates a short visual-feedback program. Researcher-refactored versions reduce mean local-task time by 29-31% in 27 simulation trials. Finally, 12 real-robot trials using operator-confirmed button contact demonstrate sim2real reuse: at a shared nominal start, simulation XML assets and simulation experience reduce mean time by 53.0% and 49.9%, respectively; real experience also transfers to two new starts. These results suggest a practical way to build general-purpose manipulation experiments around GPT-6-Astra: supply machine-readable body descriptions and synchronized demonstrations, and turn useful agent-generated feedback routines into reusable skills, while the agent adapts actions from current images. We release all task prompts, trial-level experimental data, and acquired skill implementations at https://github.com/hesd10/astra-robot-sim2real. --- ### UNMATCH: Selective Unbalanced Token-Patch Matching for Forensic Image-Claim Verification Authors: Xinjin Li, Lian Lian, Yuanzhe Yang, Yudi Xia, Calvin Chang Liu, Yeyun Xu, Yu Ma, Jinghan Cao, Yuruo Gong Published: 2026-09-24 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.31766v1 Contextual image misuse pairs an image with a misleading claim. We study image-claim correspondence in fact-checked pairs containing out-of-context reuse, visual manipulation, or both. Existing pair-based detectors often compress the two modalities into a global compatibility score or learn a highly flexible interaction module, which can obscure a decisive local mismatch. We introduce directional multiscale coverage, a compact representation that summarizes local image-claim affinity in both directions and at three spatial scales. At each scale, each direction is summarized by its mean, lower quartile, and two thresholded support ratios; the signed difference between directional means completes a nine-dimensional scale descriptor. Concatenating the three scales yields a compact local representation for a lightweight global-local classifier. Under leakage-aware three-fold, three-seed evaluation on the Snopes subset of the Fauxtography benchmark, UNMATCH achieves 69.82 Macro-F1 and 71.05 balanced accuracy, exceeding the MCOT adaptation by 2.60 and 2.16 points. A matched-reassigned intervention shows that breaking the observed pairing lowers coverage and increases both discrepancy and false-pair probability. --- ### HGPTrans: Hierarchical Graph-Pooling Transolver for Automotive Aerodynamic Drag Coefficient Prediction Authors: Bo Liu, Qiuli Luo, Lianrui Nie, Fengli Zhang, Wenjiang Wang Published: 2026-09-24 Categories: cs.CV, physics.flu-dyn arXiv: https://arxiv.org/abs/2609.31765v1 Accurate and rapid prediction of the aerodynamic drag coefficient ($C_D$) is essential for vehicle design, particularly during early-stage styling iterations where a large number of candidate geometries must be evaluated. Although computational fluid dynamics (CFD) provides reliable aerodynamic estimates, its high computational cost, typically requiring hours to days for a single configuration, limits its use in large-scale design exploration. This paper proposes HGPTrans, a hierarchical graph-pooling network with Transolver-based attention, to directly predict $C_D$ from vehicle surface meshes. Motivated by the fact that vehicle aerodynamics depends on both local geometric features and long-range interactions among spatially distant surface regions, HGPTrans integrates three complementary components. Graph isomorphism convolutions encode discriminative local geometry, physics-aware slice attention captures global interactions with linear computational complexity, and information-redundancy-aware hierarchical pooling progressively removes redundant nodes while preserving informative geometric structures. The model is trained and evaluated on the large-scale DrivAerNet and DrivAerNet++ datasets, where it achieves the lowest mean absolute error and mean squared error among the evaluated baselines. Its generalization capability is further assessed through transfer learning on a real-vehicle dataset containing both sedans and SUVs, achieving relative $L_1$ errors of 1.56% (sedans) and 2.12% (SUVs) with an inference time of approximately $0.293$ s per vehicle. This corresponds to an acceleration of several orders of magnitude relative to high-fidelity CFD while keeping the predicted drag coefficients within a few percent of the CFD reference. Ablation studies confirm each component's contribution and reveal the effects of depth and pooling ratio. --- ### SMARtCARE: Privacy-Preserving Agentic AI Systems for Bounded-Autonomy Clinical Decision Support Authors: Srini Ramaswamy, Deveeshree Nayak Published: 2026-09-24 Categories: cs.AI, cs.ET, cs.SE arXiv: https://arxiv.org/abs/2609.31763v1 Long-context clinical AI systems can miss relevant patient history when prior admissions fall outside the active reasoning context. In ICU monitoring, this can cause early vital-sign drift to appear nonspecific even when it resembles a prior deterioration pattern. SMARtCARE addresses this gap through a four-state clinical decision-support architecture: Stable, Meta-cognitive, Assisted, and Regulated (Revoked). Rather than automatically retrieving prior records, SMARtCARE uses a lossy six-channel fingerprint of the patient's prior trajectory. When current drift matches that fingerprint and the prior record is absent from context, the system raises a Meta-cognitive escalation for clinician review; full retrieval occurs only through clinician action in the Assisted state. A patient-identity guard is designed to enforce correct attribution across data loading, logging, and audit layers. Evaluation combines a synthetic Monte Carlo study that validates the state-transition logic and estimator stability, not clinical performance, with real-data runs on both the MIMIC-III and MIMIC-IV Clinical Database Demos. On MIMIC-III, one prior-pattern recurrence was identified among 14 two-admission patients; on MIMIC-IV, the same pipeline produced no fingerprint matches among 9 two-admission patients, which illustrates a key limitation of a fixed canonical pattern library. Across both runs all logged decisions were fully traceable and correctly attributed. The results support SMARtCARE as a traceable, privacy-aware mechanism for surfacing middle-context risk; they are not a clinical efficacy claim. --- ### In-Context Binding Capacity in Language Models Authors: Manas Venkata Sai Ravulapalli, Samrath Singh Chadha Published: 2026-09-24 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.30634v1 How many assignments can a language model recall before it loses track of which value belongs to which entity? We measure this limit using continuous recall curves for 12 models at or below 3B parameters and a threshold sweep over 30 open models up to 12B. On the continuous curves, the load at which recall falls halfway to chance follows $K_{50}=cN^α$, with $α=0.820$ and $R^2=0.73$. The broader sweep shows an eightfold range associated with pretraining recipe, although the continuous curves show no detectable recipe effect after controlling for scale, with few modern models in the fit. We derive why interference can lower measured capacity by reducing single-binding recall even when the load-dependent recall profile is unchanged. Direct task training also exceeds the extrapolated zero-shot law, but different measurement criteria prevent interpreting that comparison as a capacity gain. Its formation times follow a power-law form in two independent codebases, conditional on runs that succeed. Together, these results characterize capacity at the model's query interface. Bounds on joint recall and a decomposition of policy errors connect this measurement to working memory and instruction following, without treating recall as a measure of alignment. The controlled task also provides a baseline for testing whether binding limits constrain world-state tracking; the present experiments do not measure state updates or downstream transfer. --- ### Stable initialization without the CLT Authors: Simon Kuang, Kyle Chickering, Xinfan Lin Published: 2026-09-24 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.30633v1 Successful training of deep neural networks is highly dependent on the distribution of the initial weights. If the weights are too large, network training blows up; if they are too small, the model fails to learn features. Stable initialization is the optimal moderation between these two extremes. The conventional theory of random networks uses the Central Limit Theorem to control inter-neuron dependencies, which introduces distributional approximation error and coupling between layers. For networks with sine activations, we derive the uniform-phase initialization, which obviates distributional approximation and fully decouples the layers. Ours is the first work to use the sine function's periodic symmetry. Models trained with the uniform-phase initialization outperform the state of the art in neural representation tasks like image and audio fitting. We find that our untuned models are competitive with the best-tuned baselines from previous work and support $μ$P width scaling. --- ### Adapting Personalized Speech Enhancement for Low-Latency Audio-Visual Target-Speaker Extraction Authors: Rayhan Rashed, Senja Filipi, Ross Cutler Published: 2026-09-24 Categories: eess.AS, cs.CV, cs.SD arXiv: https://arxiv.org/abs/2609.30631v1 Online audio-visual target-speaker extraction aims to remove competing voices while preserving speech quality and bounding lookahead. Existing extractors are built and evaluated for separation on synthetic mixtures, leaving listening quality and meeting behavior largely untested. We introduce Audio-Visual Personalized Voice Quality Enhancement (AV-PVQE), which approaches these requirements from the other direction. We start from a personalized speech enhancement model that reconstructs a requested voice at high quality but confuses the target in 46% of two-speaker mixtures despite clean enrollment. Adding mouth features at its speaker-conditioning input and jointly fine-tuning the visual and reconstruction networks reduces this rate to 1.6%, with no future frames and 20 ms of algorithmic delay. Compared with an online autoregressive audio-visual extractor, AV-PVQE yields separation gains on two synthetic benchmarks and larger gains on recorded meetings, and keeps its advantage on excerpts with more speakers than the fine-tuning mixtures. In personalized P.835 listening tests on two meeting corpora, it improves overall quality over this extractor by 0.57 and 0.63 MOS, with similar mean rating relative to the starting model. Preservation and rejection tests show that it keeps the target intact when no competing voice is present and suppresses competing speech when the target is absent. --- ### FRESHLATENT: Channel-Aware Latent Adaptation for Resource-Constrained Embodied VLM Perception Authors: Rajat Bhattacharjya, Minwoo Kim, Arnab Sarkar, Tamoghno Das, Sing-Yao Wu, Eli Bozorgzadeh, Marco Levorato, Nikil Dutt Published: 2026-09-24 Categories: eess.SP, cs.CV, cs.DC, cs.LG, cs.RO arXiv: https://arxiv.org/abs/2609.30629v1 Mission-critical UAVs increasingly rely on split vision-language model (VLM) perception under tight onboard-resource and wireless-communication constraints. However, corruption of transmitted intermediate features creates a deployment mismatch for clean-trained split interfaces, while stronger channel-aware codecs can impose substantial onboard cost. We present FreshLatent, a lightweight channel-aware latent adapter that trains a power-normalized encoder-decoder through wireless corruption while keeping the surrounding VLM frozen. We formulate deployment around a mission-conditioned perception requirement and embedded interface cost, linking channel quality and communication budget to the operating conditions under which perception remains usable. At 0 dB and the tightest communication budget, FreshLatent improves gIoU and cIoU over clean split compression by 20.79 and 20.87 points, respectively. At the most adverse evaluated SNR (0 dB), across all three communication budgets, FreshLatent recovers 63.5-69.1% of the gIoU improvement achieved by a much heavier, range-trained feature-JSCC codec. On an NVIDIA Jetson AGX Xavier in 10-W mode, FreshLatent uses 37-40x fewer encoder parameters, 7.7-9.9x lower edge-interface latency, and 8.8-10.0x lower edge-interface energy than the heavier codec. Together, these results show that lightweight channel-aware adaptation can recover a substantial fraction of the robustness of a much larger communication interface while broadening quality-valid operation under constrained wireless conditions. --- ### OpenHail: An Event-Driven Gymnasium Environment for Electric Ride-Hailing Fleet Control Authors: Tommaso Schettini, Nicholas D. Kullman, Jorge E. Mendoza Published: 2026-09-24 Categories: cs.LG, math.OC arXiv: https://arxiv.org/abs/2609.30628v1 Machine-learning policies have attracted increasing interest for ride-hailing fleet control in recent years. Reinforcement learning, in particular, requires a structured simulation environment that specifies observations, actions, rewards, and decision epochs for training and evaluation. For electric fleets, this environment must also capture the interaction among stochastic demand, vehicle operations, and capacitated charging infrastructure. We present OpenHail, an open-source Gymnasium environment for joint control of electric ride-hailing fleets. Its fixed-size observation--action interface exposes request assignment, repositioning, and charging to a single policy. The event-driven simulator represents requests with pickup deadlines, vehicle job queues, battery dynamics, and finite-capacity charging facilities with first-in--first-out queues. A configurable decision-epoch mechanism separates internal simulator events from policy interactions, supporting event-driven, periodic, hybrid, and policy-requested control within the same operational model. The software provides seeded instances, feasible-action utilities, evaluation tools, operational metrics, and baseline policies. The source code is available at https://github.com/tommaso-schettini/openhail. --- ### Audio LLMs Know When They Can't Hear You Authors: Amirhosein Javadi, Richa Dixit, Mehrdad Farajtabar, Minsik Cho, Devang Naik, Mohammad Samragh Published: 2026-09-24 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.30625v2 Audio large language models allow users to interact with the model through speech. When an input recording is too degraded, the model may misinterpret the user's query and respond based on an incorrect transcription. In this paper, we study model-conditional transcription reliability: whether an Audio LLM can recognize when its own transcription is unreliable. We first prompt the Audio LLM to assess whether its own transcription would be reliable, and find that the model is a poor judge of its own transcription reliability: in most cases, it predicts that its transcription will be reliable. We find that existing approaches, including speech quality predictors, audio LLM generation uncertainty, and transcript-conditioned WER estimation, provide limited signals for detecting transcription failures. In contrast, we discover that transcription reliability is strongly represented in the model's audio-encoder representations. Based on this observation, we devise a lightweight reliability predictor that operates on representations extracted by the frozen audio encoder and predicts the reliability class before generation. The reliability predictor can trigger a clarification request from the user when their voice query is predicted to be unreliable, while allowing reliable queries to proceed without modifying the underlying Audio LLM. Our predictor achieves 81.10% in-domain and 78.09% cross-domain macro-F1 scores, outperforming the strongest baselines by 10.33 and 11.93 points, respectively. Finally, we show that reliability labels can transfer across Audio LLM families, and that transfer performance is closely related to the alignment of their model-specific reliability boundaries. --- ### Subjects, Not Authors: The Authorship Hazard in Agentic Dataspaces Authors: Seungho Lee, Changbin Lee Published: 2026-09-24 Categories: cs.CR, cs.AI, cs.DB, cs.MA arXiv: https://arxiv.org/abs/2609.30614v1 Dataspace connectors decide whether a transfer may occur, not what the transferred value contains, tolerable for contracted applications, not for LLM agents that compose tool calls and spawn sub-agents. Research on agents that generate governance artifacts evaluates output quality; who may authorize an artifact for use falls between that literature and the governance literature, and neither owns it. A published policy is what a dataspace's decision point enforces, so publication is a governance event, and an agent that is both policy subject and policy author writes the norms that bind it. We name this the authorship hazard and state one principle: an agent is a subject of the governance plane, never an author of it. Its authorization channel to publication is closed by construction; its influence channel, drafting what humans approve, is treated as an enforcement problem. On a frozen corpus of agent drafts, publishing without approval reverses 80 authorization decisions, most through drafts that change only a field's sensitivity classification and no policy text; a classifier that reads the policy diff misses every such draft, necessarily. Treating classification as authorship routes them all to review; the registry-held classification this requires is designed and modelled here, not yet implemented in the prototype. At the execution boundary, protected fields reach the model in 105 of 105 cases under prompt-stated duties and in 0 of 105 when the ODRL duty is compiled into an invocation-time tool-call constraint, but where the value is not confined to a named field the compiled condition exposes it in 7 of 7. A centrally provisioned approval pool does not scale to the participant volume that motivates the problem. --- ### MedTokenBudget: Lesion-Preserving Token Routing for Dermoscopic Image Classification Authors: Zhexiang Li Published: 2026-09-24 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.30613v1 Dermoscopy classifiers built on Vision Transformers process all image patches uniformly, although diagnostic evidence is concentrated in the lesion region. Existing token pruning methods reduce tokens using generic saliency or similarity signals, but rarely ask whether the retained subset still contains the lesion. This paper introduces MedTokenBudget, a supervised post-backbone token routing framework that learns to construct compact lesion-enriched representations when auxiliary lesion masks are available. Its Lesion-Aware Token Scoring (LATS) module fuses attention entropy, feature norm, and local feature contrast through a learned scorer, then routes the top-$K$ patches under a target budget. LATS is trained with budget curriculum learning, diversity regularization, attention distillation, and lesion-mask supervision. The trained router is evaluated with a lesion retention rate that directly measures how much ground-truth lesion evidence survives the token budget. On ISIC 2019, mask-supervised LATS consistently outperforms Random and ToMe at headline budgets while retaining substantially more lesion patches. Code is provided for reproducibility, and complete tabulated results are included in the supplementary material. --- ### Epstein Files Engine: Agentic Search for Investigative Journalism Authors: Duy K. Nguyen, Teresa Mondría Terol, Dylan Freedman, Zach Seward Published: 2026-09-24 Categories: cs.HC, cs.CL, cs.CY, cs.IR arXiv: https://arxiv.org/abs/2609.30611v1 On Jan. 30, 2026, the U.S. Department of Justice released a mixed-media collection concerning Jeffrey Epstein, including about three million pages of PDFs. We describe the Epstein Files Engine, an A.I. agent The New York Times deployed to investigate the files. The Engine translated reporter questions into Google BigQuery SQL queries across three corpora: Epstein-related releases, the Times's archive and external, Epstein-related news headlines. It used an LLM to plan queries and returned citation-rich answers a reporter could verify and trust. More than 100 journalists used the Engine, and it contributed to at least 20 published stories. We report how reporters queried it and describe Diff, our text-and-visual duplicate matching method that amplified novelty signals and allowed the Engine to surface genuinely new information. We argue that newsroom agents serve newsrooms best not as autonomous writers, but as interfaces to source material and institutional knowledge. --- ### MVAgent: Multi-Agent Video Generation via Consistent Condition Construction and Shot-Level Policy Optimization Authors: Xiangyu Kong, Wenjie Zhou, Fengping Tian, Lihua Fang, Haoqin Sun, Chenyang Lyu, Longyue Wang, Weihua Luo Published: 2026-09-24 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.30609v1 Multi-shot agentic video generation requires consistent character appearance, stable spatial layout across camera angles, and continuous character state between shots. When every shot is a separate request to a frozen generator, repeated text does not determine appearance, layout or state. We therefore recast the problem as condition construction and present MVAgent, a multi-agent pipeline whose agents collaborate through typed conditioning inputs. Because an environment image shows one viewpoint, a Spatial Grounding agent samples views from generated camera-traversal clips and anchors each shot to the view matching its framing. As generated shots drift from the plan, an Observer records how each shot ends in a continuity memory, from which a Transition agent builds character action and spatial references for the next shot. An Orchestrator composes these inputs into each request. Since a request reveals its effect only after rendering, we train it by agentic reinforcement learning with Trunk-GDPO, which compares rendered candidates at every shot rather than once per video and continues the best as the trunk. With generator and judges frozen, MVAgent attains the highest cross-shot consistency and narrative-planning quality among the compared methods on ViMax-Bench and is preferred over the strongest agentic baseline in human evaluation. --- ### Probabilistic Robustness-driven Universal Adversarial Perturbations with Explainability against Deep Reinforcement Learning-based Intrusion Detection System Authors: Hongsen Zhang, Lu Zhang, Mingjing Xu, Yi Zhang, Gregory Epiphaniou, Carsten Maple Published: 2026-09-24 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.30605v1 Deep reinforcement learning (DRL) enables adaptive intrusion detection in dynamic network environments but also exposes intrusion detection systems (IDS) to adversarial threats such as universal adversarial perturbations (UAPs), which apply a single input-agnostic perturbation to degrade detection performance across traffic. Probabilistic Robustness (PR), as a post-hoc evaluation metric, provides a principled, population-level measure of adversarial impact that conceptually aligns with the universality objective of UAPs, i.e., PR quantifies the prevalence of misclassification in the input space, making it a natural signal for guiding UAP generation. Hence, we propose PR-based UAP, which represents the first integration of an explicit PR-driven objective into generating UAPs against DRL-based IDS. Building on this formulation, we introduce PX-UAP, which leverages explainable artificial intelligence (XAI) to guide perturbation shaping under realistic domain constraints, and provides a rigorous theoretical analysis of its design. Extensive experiments demonstrate that PX-UAP consistently outperforms state-of-the-art UAP methods in attack effectiveness. --- ### The Hard Part Comes After Search: Benchmarking Web Agents on Synthesizing, Organizing, and Displaying Knowledge Authors: Alexander Gill, Md Farhan Ishmam, Xuyen Nguyen, Neha Bhat, Parker Henry DeYoung, Fateme Hashemi Chaleshtori, Nathan Stringham, Kenneth Marino, Ana Marasović Published: 2026-09-24 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2609.30604v1 Existing computer-use agent benchmarks do not fully evaluate agents acting as assistants. A useful assistant retrieves information across complex, multi-step workflows, synthesizes it into artifacts (documents, presentations, spreadsheets), and navigates program interfaces to produce a coherent final product. Such workflows demand reasoning and synthesis, decomposition of complex tasks, as well as visual and spatial understanding. To study agents on workflows like these, we introduce KNOWS, a benchmark of open-ended, complex, browser-based tasks that jointly evaluate these capabilities, with each task culminating in a produced artifact. To write tasks, we develop a task design rubric and a protocol for ensuring that tasks meet the requirements. Each task is paired with an evaluator, a program that combines deterministic checks with LLM judgments to balance the richness, reliability, and automation tradeoff inherent to agent evaluation. We evaluate and analyze frontier computer-use agents and browser-based harnesses. They achieve moderate scores on partial-success metrics, but the best performer fully succeeds in fewer than 3% of our complex, long-horizon tasks. Failures on visual steps render the resulting artifacts unusable, even when agents complete more than 50% of other evaluation steps. Our results expose limitations of current agents acting as end-to-end assistants, and call for progress on tool use, visual understanding, and long-horizon reasoning. --- ### Action Forcing: Training World Models on Unsupervised Video by Recovering Underlying Egomotion Bases Authors: Ashish Sundar, Tiankuo Hou, Zhong Fan, Chunbo Luo, Xiaoyang Wang Published: 2026-09-24 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.30595v1 Synchronised action annotations are needed to train controllable world models and these datasets remain elusive. Existing approaches make use of instrumented platforms with calibrated sensors, costly manual annotation, or latent-action models which lack grounding. We instead turn ordinary unlabelled video into action-supervised training data by recovering (without training) a data-derived egomotion basis. We track pixel displacements across frames and exploit the recurring coherent structure induced by egomotion to obtain grounded control signals directly. Using a method as simple as principal components analysis perform this, we find that the leading components provide signed, scalable, and composable throttle--yaw controls, although the method can recover only motion axes represented in the data. To prevent a high-capacity video DiT from exploiting pixel-level supervision, an online latent critic distils a frozen decoder--tracker--PCA (Principal Components Analysis) teacher without backpropagating through the decoder or tracker. Finally we critique the use of video generation metrics to evaluate WMs and introduce an example of an alternative, reference-free evaluation method. We measure \textit{controllability}, \textit{plausibility}, \textit{conjuring} (creating objects out of thin air) and \textit{geometric integrity}, revealing failures that conventional video metrics miss. We show that most baselines follow familiar action directions but struggle to reverse or remain stationary. Our model handles both while retaining compositional control and generation quality. Despite backwards actions being less than $1\%$ of our training data, we find that the model learns to reverse, scale its response linearly, and compose throttle with steering, all simply by learning through a grounded action space. --- ### QSV: Quat-Sphere-Vision for Coupled Quaternion Attention on Spherical Lattices Authors: Nicholas Foley, Devin Marinelli, Donny Moore, Diego Enriquez, Amanda Fernandez Published: 2026-09-24 Categories: cs.LG, cs.CV arXiv: https://arxiv.org/abs/2609.30592v1 In standard attention, three separately learned projections decide how strongly a token attends to each neighbor ($W_Q$, $W_K$) and how the attended features are transformed before aggregation ($W_V$). We study Quat-Sphere-Vision (QSV), a sparse spherical vision model that replaces this projection triple with a single learned unit quaternion per token: the relative quaternion $r_{ij} = q_i^{*} \otimes q_j$ supplies both the attention logit $\operatorname{Re}(r_{ij})$ and a sandwich-product feature transport $x \mapsto r_{ij} \otimes x \otimes r_{ij}^{*}$, with messages passed over sparse kNN graphs on concentric Fibonacci spheres. Ablations that change only the targeted component show the two roles to be asymmetric. Removing the transport reduces test accuracy by about four percentage points on CIFAR-10 and CIFAR-100 (single runs per CIFAR-100 variant), while replacing the learned attention weights with uniform averaging leaves it essentially unchanged. Parameter-matched controls then remove the geometry itself: standard attention on the same graph exceeds QSV (mean $87.3\%$ vs. $85.9\%$), and the same model on a flat 2D lattice reaches $91.1\%$, within $2.1$ points of a ResNet-20 trained under the same pipeline (single run). In the coupled kernel, nearly all of the learned pairwise computation resides in the transport channel. --- ### Encryptability As a Coordinate Choice: Depth-One Homomorphic Federated Learning of Quantum Neural Networks Authors: Marcel Mordarski, Nathan Mani, Arshad Patel, William Knottenbelt, Roberto Bondesan Published: 2026-09-24 Categories: quant-ph, cs.CR, cs.DC, cs.LG arXiv: https://arxiv.org/abs/2609.30581v1 Encrypted training relies on keeping server-side updates low-degree. This constraint traditionally excludes models whose weights inhabit a compact Lie group (notably variational quantum circuits, where every trainable weight is an $\mathrm{SU(2)}$ rotation). Expressed in Euler angles or discrete alphabets, these updates appear transcendental, historically demanding prohibitive costs: one client--server round per gate, or upwards of $25{,}000$ operations per weight. This penalty is strictly an artefact of coordinates. In the unit-quaternion (spin) chart, group composition is exactly bilinear (degree two, with coefficients in $\{-1,0,+1\}$). Consequently, encrypted rotation updates cost one multiplicative level and federated averaging costs zero in any levelled homomorphic scheme, completely eliminating bootstrapping. This implementation-independent algebraic property is confirmed across two cryptographic backends, introducing only $0.0$ and $-2.0\times10^{-12}$ rad of aggregation error. Leveraging this reduction yields a non-interactive protocol for encrypted federated training of hybrid quantum--classical networks. It includes correctness proofs for aggregation and sign handling, plus a compilation lemma proving parameterised entanglers add only constant-factor overhead without altering the depth class. Empirically, a paired five-seed study confirms zero measurable utility tax ($Δ=+9\times10^{-6}$ MSE, $p=0.92$), and a noise-budget ablation falsifies the hypothesis that encryption noise regularises. These convergence trends replicate across datasets and scale to $20$ clients. Finally, hardware validation on a $156$-qubit processor achieves $0.9918$ fidelity against a $0.99957$ unencrypted control. --- ### Energy-efficient operation of neural operators for virtual sensing Authors: Jason Yoo, Samrendra Roy, Souvik Chakraborty, Syed Bahauddin Alam Published: 2026-09-24 Categories: cs.LG, cs.PF arXiv: https://arxiv.org/abs/2609.30580v1 Virtual sensing repeatedly reconstructs physical fields from changing observations, often on a fixed geometry. We investigate how shared spatial computation reduces the energy of these updates while retaining the selected checkpoint and its evaluated predictions. In a heat-exchanger service, standard compiler freezing and explicit trunk reuse give similar operating energy reductions relative to graph replay: approximately 1% at one request per second and 20% at forty requests per second. In 15 W mode with fixed clocks, reuse with graph replay completes the same request sequence with 22.0 to 22.5% less energy than eager execution, including preparation and waiting. DeepONet and Fourier neural operator (FNO) controls distinguish the effects of reusable arithmetic and launch overhead. Preparation, artifact construction, and worker replacement add costs outside repeated inference. These results connect operator structure to operating energy and show how update frequency and execution lifetime govern the benefit of computation reuse in physical-field virtual sensing. --- ### Reinforcement Learning of Communication in a Mesh of Small Language Models Authors: Mehmet Kerem Turkcan Published: 2026-09-24 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.30578v1 Language models gain accuracy from more compute at test time, but majority voting over independent samples saturates: as samples grow, the vote converges to the model's most frequent answer. Communication can add what sampling cannot: an agent that solves a problem can pass the key step to the others. We present TalkMesh, a decentralized mesh of small language model agents that learns when and what to communicate. Each agent samples a proposal and scores it with a trained confidence head. The most confident agent broadcasts a hint; agents below a confidence threshold revise, keeping each revision that outscores its proposal. Gossip consensus approximates the vote weighted by confidence without a coordinator. A talk policy, trained with group relative policy optimization on the change in correctness after revision, writes hints and revisions. With three agents, which together generate at most six outputs, the mesh reaches the accuracy of majority voting over 32 samples with each of three models. Trained with at most 8 agents and evaluated with 32, it raises accuracy from 0.568 under self-consistency to 0.705 (Qwen3.5-0.8B, GSM8K) and from 0.492 to 0.722 (SmolLM3-3B, MATH-500). When 4 of 8 agents collude on a wrong answer with fabricated confidence and poisoned hints, majority vote accuracy falls to 0.000 (Qwen3.5-0.8B, GSM8K). A defended mesh, whose agents rescore solutions with their own confidence heads, retains 0.507. Across reasoning, embodied coordination, and traffic signal control, messages improve a decision when the acting agent cannot observe the information it requires and another agent can send it. --- ### T-RoPE: Time-Aware Rotary Position Embedding for Sequential Recommendation Authors: Yang Liu, Noel Loo, Ali Khanafer, Shuying Sun, Akshay Soni, Zhong Wu, Linjun Yang Published: 2026-09-24 Categories: cs.AI, cs.IR, cs.LG arXiv: https://arxiv.org/abs/2609.30576v1 Large-scale recommenders increasingly adopt the sequential generative recipe behind large language models, bringing the Transformer into recommendation along with design choices made for text, including Rotary Position Embedding (RoPE). In language models, RoPE encodes token indices for relative position reasoning, but in recommendation, an interaction index records only event order, saying nothing about elapsed time, behavioral cycles across scales, or calendar phase. We revisit this choice and propose T-RoPE, a time-aware RoPE for sequential generative recommendation that replaces index-only rotation with timestamp-based angles, learnable temporal coefficients, multiscale frequency banks, shifted query alignment, and non-stationary key rotation. We prove that standard RoPE, even on timestamps, remains time-translation invariant and cannot distinguish seasonal contexts, and that T-RoPE breaks this invariance while preserving the RoPE interface. Across five public benchmarks, T-RoPE achieves the best result on every metric on every dataset, improving over the strongest baseline by 78--130\% in HR@10 on the sparse PixelRec data and 8--12\% across metrics on Amazon Books. On an industrial-scale e-commerce dataset with more than 6B interactions, it improves every metric over the HSTU + Time RAB backbone by 13--82\%, with ablations attributing the largest gains to multiscale frequencies ($+56\%$ NDCG@50) and non-stationary keys ($+4\%$). An online A/B test in the Shop app yields positive lifts in conversion rate ($+0.33\%$) and order count ($+0.63\%$). We also provide forward and backward algorithms whose added cost is linear in sequence length and head dimension, keeping time-aware RoPE practical for large generative recommenders. --- ### Entropy Regularization: A Free Correction to Cross-Entropy for Verified Demonstrations Authors: Mihir Dhanakshirur, Adam Ousherovitch, Ambuj Tewari Published: 2026-09-24 Categories: cs.LG, stat.ML arXiv: https://arxiv.org/abs/2609.30572v1 Large language models are often post-trained on expert demonstrations using cross-entropy (CE), even when the downstream objective is not to imitate the demonstrated solution but to produce any output accepted by a verifier. This mismatch is seen in verifiable domains with multiple correct solutions, such as mathematical reasoning and code generation, where training data may contain only one expert solution per problem. We show that minimizing cross-entropy can be misaligned with minimizing verifier risk; two policies can assign identical likelihood to the observed demonstrations while placing different probability mass on incorrect outputs. This is formalized through a learning-theoretic counterexample in which CE minimization selects a suboptimal policy. We identify that controlling the support of the learned policy can solve this problem by preventing probability mass from spreading to unsupported outputs. Since support size is non-differentiable and computationally intractable, we propose entropy-regularized cross-entropy (ER-CE), using token-level Shannon entropy as a tractable proxy. Finally, across mathematical reasoning and code-generation benchmarks, we find that entropy-regularized training consistently improves verifier accuracy over standard cross-entropy. Our results identify a simple failure mode of imitation-based post-training in verifiable tasks and provide a practical objective that is better aligned with producing correct outputs. --- ### HARDEN: Constrained Evolutionary Search for Harder, Answer-Preserving Evaluation Cases Authors: Aditya Kumaran, Rahul Singhal, Karime Maamari, Amine Mhedhbi, Pradyumna Tambwekar Published: 2026-09-24 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.30571v1 Language models are often evaluated on curated benchmarks that underrepresent the complexity of enterprise deployments. We introduce HARDEN, a constrained evolutionary search method to adapt the input of existing evaluation cases into more challenging variants while keeping their expected outputs fixed. HARDEN searches along generated domain-specific complexity axes while enforcing feasibility constraints such as preserving task semantics, realism, and execution validity. Across FinQA, PubMedQA, and ContractNLI and three Qwen3.5 model scales (35B-A3B, 122B-A10B, and 397B-A17B), HARDEN reduces task-model accuracy by 22.7% on average and by up to 49.9% relative to single-pass baselines using the same feasibility checks. These results show that evolutionary search can produce substantially harder valid evaluation cases. --- ### Atelier: Learning Local Self-Supervised Features for CryoEM Volumes via Hypernetworks Authors: Phillip Lo, Sudarshan Babu, Dari Kimanius, Aly A. Khan Published: 2026-09-24 Categories: cs.AI, stat.ML arXiv: https://arxiv.org/abs/2609.30569v1 CryoEM map interpretation requires features that are spatially localized, consistent across samples, and informative across spatial scales. Most deep learning methods for map annotation extract features from fixed voxel grids. However, implicit neural representations (INRs) are able to model volumetric data as scale-agnostic, coordinate-conditioned functions. INRs are therefore attractive for cryoEM, but fitting a separate INR for each map is too expensive for large-scale feature extraction and produces representations that are not aligned across samples. We introduce Atelier, a self-supervised framework that amortizes INR fitting for reconstructed cryoEM maps. Pretrained on 5,439 Electron Microscopy Data Bank maps, Atelier is a transformer-based hypernetwork that generates high-fidelity reconstructions across a wide range of protein structures, including large multi-subunit assemblies. Beyond reconstruction, the INR generated by the pretrained transformer exposes a continuous, local feature field through its intermediate activations at any spatial query point, a property that voxel grid and patch-tokenizer architectures do not naturally provide. Used as auxiliary channels to a 3D nested U-Net annotation head trained from scratch, these coordinate-conditioned features improve performance on eight voxel-level property prediction tasks over a volume-only baseline. Our results demonstrate that amortized implicit neural representations are an effective primitive for geometry-aware analysis of cryoEM data. --- ### Atlases Are Already Inside: Recovering Population Templates from Pretrained Diffusion Models Authors: Jian Shi, John Femiani, Peter Wonka Published: 2026-09-24 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.30566v1 We present a new inference-time sampler for diffusion models that gives a pretrained model a capability it was never trained for: constructing the atlas of the population it synthesizes. The sampler converges from every random seed to the population's central anatomy, which we call the \emph{intrinsic atlas}. The advantage is threefold. (1) It requires no retraining. A diffusion model that has already learned a coherent population, including the released ones, yields its atlas in a single inference pass without involving deformable registration. (2) It applies to multiple domains, such as brain MRI, chest X-ray, faces, and 3D shapes. (3) It extends to subpopulations. One age-conditioned model gives an atlas at any age in its training range, and the resulting family reproduces the CSF expansion of healthy aging. Evaluated as a registration target, the intrinsic atlas is best or second-best on every dataset against classical and learned templates, and the most central template on held-out brain MRI cohorts. Atlas construction can be reframed as a byproduct of generative modeling: a diffusion model is a learned representation of population structure, and the atlas is what it already contains. --- ### High-dimensional Gaussian Graphical Model Testing for Long-Memory Time Series Authors: Percy S. Zhai, Ping-Shou Zhong, Wei Biao Wu Published: 2026-09-24 Categories: stat.ME, math.ST, stat.ML arXiv: https://arxiv.org/abs/2609.30565v1 Many real-world high-dimensional time series exhibit long-memory, but Gaussian graphical model testing in this regime remains understudied. We develop a direct, data-adaptive test statistic for assessing conditional independence in the graph structure of stationary Gaussian time series. We establish a finite-sample, Berry--Esseen type Gaussian approximation bound for the statistic, which applies to both short-memory and long-memory time series. The testing procedure is fully data-adaptive using block bootstrap method, on which we provide a finite-sample validity result including in the ultra-high-dimensional scenario, and can be extended to comparing graphical structures in two-sample tests. We also develop a consistency-empowered correction to the statistic and show that such tests attain asymptotic consistency in both size and power. Our proposed method is applied to a real-world fMRI data to understand functional connectivities within brain in different periods. --- ### Thinking Less to Simulate Better: Intuitive Prompting Improves LLM Agents Simulating Individual Social Media Reactions, Including Unfamiliar Content Authors: Ljubisa Bojic, Tijana Stanic, Joerg Matthes, Agariadne Dwinggo Samala, Bojana Dinic, Jue Wang Published: 2026-09-24 Categories: cs.AI, cs.CL, cs.HC, cs.MA, cs.SI arXiv: https://arxiv.org/abs/2609.30563v1 Platform policies are increasingly tested on artificial users, making agent fidelity important. Yet convincing fake profiles could also manipulate perceived public opinion before elections. Validation has concentrated on agreement with human behaviour and has paid little attention to whether an agent behaves in line with the profile it was given. The present study profiled eight Serbian participants through a questionnaire, a deep interview, and a written self-presentation, recorded their reactions to sixty-eight social media posts, and asked four language models to predict those reactions under five prompt conditions varying profile content and instruction style. Attitudinal content improved prediction over demographic backstories by a wide margin. Agents matched their stated profiles more closely than participants matched their own survey answers, and consistency proved unrelated to fidelity once profile information was present. Instructing models to respond intuitively and immediately rather than analytically gave the highest fidelity of any condition and cut the compression of individual differences from seven times the human level to three. The advantage held on posts about topics the questionnaire never raised, where that condition reached the highest fidelity of any setup and beat a crowd baseline by a wide margin, which suggests that agents prompted this way could serve as general-purpose simulated users rather than specialists on the topics they were profiled for. Results may bear implications for the development of language models, because intuition-based setups appear better suited to some tasks than reasoning-based ones. --- ### Probing Stability-Plasticity Tradeoffs in Agent Memory through Cognitive Experimental Paradigms Authors: Jiaqi Ding, Guorong Wu Published: 2026-09-24 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.30558v1 Agent memory systems are increasingly used to maintain long-term user preferences, task states and evolving facts, but current evaluations often collapse memory behavior into final-answer accuracy. We introduce MemProbe, a cognitive-science-inspired framework for diagnosing stability-plasticity tradeoffs in agent memory. The framework is motivated by a core insight from cognitive memory research: memory is reconstructive and shaped by interference, source reliability, reinforcement, and reactivation. MemProbe turns this insight into four reusable experimental paradigms (interference, misinformation, consolidation strength, and reconsolidation window) that manipulate when a memory should be updated, preserved, or treated as uncertain. It further decomposes correctness into behavioral profiles that reveal how systems update, preserve, attribute, and temporally organize information. We instantiate these paradigms in a 56-episode diagnostic suite and evaluate six incremental memory systems under a unified protocol. Results show that systems with similar aggregate scores exhibit distinct behavioral profiles. MemProbe provides such a diagnostic lens, turning aggregate performance into interpretable profiles of memory maintenance over time. Code is available at https://github.com/jq-ding/MemProbe. --- ### Auditing Latent-Space Monitors for Autonomous Driving Authors: Nikhil Kamalkumar Advani, Vishwajeet Shivaji Hogale, Saurav Kumar Published: 2026-09-24 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2609.30557v1 Runtime failure monitors can use a model's internal representations to anticipate failures. We audit this monitoring strategy across two autonomous-driving tasks: online vectorized map generation with LaneSegNet and end-to-end planning with VAD. We find that frame-level errors are predictable at inference in both tasks. For LaneSegNet, a supervised latent probe reaches Area Under the Receiver Operating Characteristic curve (AUROC) 0.780 for high Chamfer error; to our knowledge, this is the first post-hoc frame-level failure monitor for online vectorized map generation. For VAD, a supervised planning-latent probe reaches AUROC 0.868 for mean-ADE failure. Our audit shows that internal access is not necessary for strong failure prediction. A monitor using only LaneSegNet's prediction outputs reaches AUROC 0.825, while for VAD, ego state, driving command, and the planner's predicted trajectory reach 0.924 on the same mean-ADE endpoint. Adding latent features to either baseline yields no statistically resolved improvement. This observation persists across a broad suite of planning failure endpoints, including endpoints whose labels depend on geometry unavailable to the non-latent baseline. Thus, predicting failure from an internal representation does not establish that the representation provides useful information beyond observable inputs and outputs. We propose an evaluation protocol for testing the incremental value of latent access and release our per-frame failure endpoint labels. --- ### Dynamic Regret in Online Convex Optimization with Indicator Switching Costs Authors: Naram Mhaisen, George Iosifidis Published: 2026-09-24 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.30556v1 We study dynamic regret in online convex optimization with an \emph{indicator switching cost}: a fixed penalty incurred whenever two consecutive decisions differ. This captures startup overheads such as server activation, model deployment, and cache updates, and on a bounded domain it recovers norm-based movement costs as a special case. Existing guarantees for indicator costs handle only static comparators. We show that a direct extension of these techniques to dynamic regret provably fails, motivating a different approach. We propose a meta-learning framework: a set of randomized lazy FTRL base learners restarted at dyadic time scales, aggregated by a movement-aware master that mixes their proposal densities and samples actions via maximal coupling of consecutive mixtures. The resulting algorithm satisfies, in expectation, $\mathcal{R}^{\mathbf{1}}_T \le \tilde{\mathcal{O}}(\min\{\sqrt{T(S_T{+}1)},T^{2/3}(P_T+1)^{1/3}\})$, where $\mathcal{R}^{\mathbf{1}}_T$ is the dynamic regret plus the cumulative indicator switching cost, $S_T$ counts comparator switches, and $P_T$ is the comparator path length. The bound holds simultaneously for all sequences and requires no prior knowledge of $S_T$ or $P_T$: it is minimax-optimal (up to logarithmic factors) for tracking piecewise-constant comparators, and also captures frequently moving comparators with small total path length. --- ### Proportional Representation in Temporal Voting with Ranked Preferences Authors: Noam Hazon, Leora Schmerler, Nicholas Teh Published: 2026-09-24 Categories: cs.GT, cs.AI arXiv: https://arxiv.org/abs/2609.30555v1 We study proportional representation in temporal voting, where one candidate is selected in each round. While prior work has focused on approval ballots, we consider ranked preferences, which may change over time. A natural approach treats each voter's top candidates as approved, but the right cutoff may differ across voters and rounds. We therefore require proportionality to hold for every admissible choice of cutoffs, whether fixed and common, common but varying across rounds, or set individually for each voter in each round. Combining these interpretations with temporal versions of justified representation (JR), proportional JR (PJR), extended JR (EJR), and proportionality for solid coalitions (PSC) gives us a hierarchy of axioms. We ask which of these axioms can be guaranteed, and with how much knowledge of the future. Unlike with approval ballots, no version of EJR can be guaranteed, and for the other axioms, flexibility in the cutoffs comes at a price. With a fixed common cutoff, JR, PJR, and PSC can be guaranteed, but only by rules that see all preferences in advance. Once the cutoff may vary across rounds, even such rules cannot guarantee JR or PSC for groups that agree in only some rounds. For groups that agree in every round, however, knowing only the number of rounds suffices for PJR in polynomial time, and PSC needs no knowledge of the future at all. Under individual cutoffs, no version of JR or PJR can be guaranteed, yet a rule as simple as serial dictatorship achieves PJR up to an additive loss that no rule can improve on, however much it knows. Natural preference restrictions restore exact guarantees. Finally, we show that checking our axioms is often coNP-complete; but perhaps surprisingly, a stronger axiom can be easier to check. --- ### Rank-Reliable Teacher-Guided Fitness Approximation for Expensive Evolutionary Optimization: A TinyML Architecture Search Study Authors: Soumen Garai, Suman Samui Published: 2026-09-24 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.30553v1 Expensive evolutionary search does not always need an exact fitness estimate for every candidate. It often needs a reliable answer to a simpler question: which candidate is better? We address this need through Teacher-Guided Learning NSGA-II (TGL-NSGA-II), a low-fidelity framework for constrained Tiny Machine Learning (TinyML) neural architecture search. A pretrained teacher organizes samples into strata defined jointly by difficulty and class. Each candidate then undergoes KD-Lite, a short and capped knowledge-distillation procedure on a compact training set, before being scored on a separate stratified evaluation set. This teacher-guided score is fused with a Gaussian-process surrogate to select candidates for full evaluation. For a fixed candidate population, we analyse evaluation variance, score concentration, pairwise rank inversion, expected Kendall-$τ$, first-front identification, and hypervolume perturbation. We also derive a variance-aware fusion weight and a capacity-adaptive distillation rule. On keyword spotting and bird-call classification, the measured Kendall-$τ$ values are 0.74 and 0.62, exceeding the corresponding predicted lower bounds of 0.60 and 0.46. Joint stratification reduces proxy-score variance by 41% relative to random evaluation. Selective teacher mismatch, in contrast, increases differential bias and reduces Kendall-$τ$ to 0.41. Under a constrained evaluation budget, TGL-NSGA-II achieves the largest mean hypervolume and smallest generational distance on keyword spotting, records the lowest mean false-positive rate on BirdCLEF, and runs 2.2x faster than full NSGA-II. These guarantees apply to population-level low-fidelity evaluation and do not establish convergence of the complete evolutionary trajectory. --- ### Benchy: towards a universal language for task-oriented AI benchmarks Authors: Francis F Daniel, Mauro Ibañez, Francis Perelman, Marian Basti Published: 2026-09-24 Categories: cs.AI, stat.ME arXiv: https://arxiv.org/abs/2609.30550v1 Benchy is a semantic language and execution engine for benchmarking AI programs. A benchmark is completely specified by a program, a scoring function, and a dataset, B=(P,S,D), and is separate from the AI-system taking it; a run binds the two, R=(B,AI). Benchmarks are authored as canonical YAML in which each semantic concept has one valid syntax, classified by a shared task/domain/language ontology, and deterministically compiled into a canonical JSON intermediate representation that the engine executes. Compilation changes representation, not meaning: it does not repair invalid definitions or inject hidden defaults. Programs use fixed schemas of named input and output fields, the leaf output fields are the scoring dimensions, and the engine exposes one universal runtime contract --- a named-field input object in, a named-field output object out --- to which external AI-systems adapt at the boundary, so integration mechanics never propagate into benchmark semantics. This paper gives the semantic object model, the ontology and task-to-program validation rule, the scoring and failure semantics, the compilation and execution architecture, and the scope of the current language. An appendix fixes the normative engineering contract for the first engine implementation. --- ### REALMS: An AI-Assistant Conversational System for Real-Time Exact Audience Sizing over High-Dimensional Nested Profiles Authors: Haixu Ma, Aditya Bansal, Shubham Lohiya, Sumit Ranjan Published: 2026-09-24 Categories: cs.CL, cs.IR arXiv: https://arxiv.org/abs/2609.30547v1 Audience sizing is a critical component of digital marketing. It enables precise resource allocation, campaign planning, and performance optimization. Traditional approaches using skeleton audiences, sampling, or predictive modeling suffer from significant delays, estimation errors, and poor scalability over high-dimensional profile data. We present REALMS (Real-time Exact Audience sizing via LLM-based Multi-attribute Search), a conversational system for exact audience sizing deployed in production on an enterprise customer data platform. REALMS enables marketers to query massive profile stores with millions of profiles and thousands of attributes using natural language and receive precise counts in seconds. The system introduces three key components: (1) a categorical attribute retrieval mechanism using embedding-based vector search to dynamically identify relevant schema attributes without manual configuration; (2) an LLM-powered NL2SQL pipeline with template-based in-context learning for accurate query generation over complex nested schemas; and (3) schema standardization enabling industry-agnostic deployment across diverse enterprise environments. Evaluation on real enterprise data demonstrates strong recall for attribute retrieval, high SQL execution accuracy, and low latency, which enables real-time interactive audience insights where prior methods required hours. --- ### Convergence guarantees for Muon: New parameter regimes and generalizations Authors: Arthur C. B. de Oliveira, Dhruv D. Jatkar, Guilherme S. Vicinansa, Eduardo D. Sontag Published: 2026-09-24 Categories: math.NA, cs.AI, math.OC arXiv: https://arxiv.org/abs/2609.30546v1 In this paper, we establish the first asymptotic convergence guarantees for the Muon algorithm through a more accurate proxy for the Newton-Schultz iteration than the typical matrix sign function. We prove that, for appropriate choices of hyperparameters, the iterates satisfy $\lim_{k\to\infty}\|\nabla f(x_k)\|=0$, and, under a global Polyak-Łojasiewicz condition, that the sequence of function values converges linearly. The key insight is that the regularization, implicit in Muon's Newton-Schulz implementation, induces a bounded preconditioner, exposing Muon as a \emph{preconditioned Polyak heavy-ball} method and enabling a classical Lyapunov analysis. This observation naturally motivates applying the same preconditioning structure to the Nesterov gradient evaluation. We formalize this idea by introducing \emph{Muesterov}, a Nesterov-based variant of Muon, and prove that it enjoys the same convergence guarantees, extending the theoretical framework beyond the heavy-ball setting. Numerical experiments on a scalar cross-entropy problem corroborate the theory and illuminate the joint role of the learning rate and the Newton-Schulz regularizer in controlling convergence. Preliminary numerical simulations training the nanoGPT dataset provide intuition regarding the relevance of the observations in this paper to practical applications. --- ### GyroNovo: Error-Guided Fragment Imputation with Mass-Aware Attention for \textit{De Novo} Peptide Sequencing Authors: Abdellah El Mekki, Laks V. S. Lakshmanan, Muhammad Abdul-Mageed Published: 2026-09-24 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.30542v1 De novo peptide sequencing from tandem mass spectra is essential for identifying peptides without relying on reference databases. Despite advances in deep learning, accurate sequencing remains challenging because experimental spectra are often sparse, noisy, and incomplete, leaving informative b- and y-ion fragments unobserved. Existing methods attempt to recover this missing evidence via latent-space imputation before autoregressive decoding. However, they typically treat imputation as a fixed reconstruction task, without considering which missing fragments are most relevant to decoder errors. Moreover, existing peak representations do not explicitly model mass differences between peaks, despite their fundamental importance. We introduce GyroNovo, a framework with two main contributions. First, we use decoder errors observed during training to adapt the imputation objective, prioritizing fragments associated with frequent decoding errors. We further use the decoder error distribution to construct easy and hard augmented views of each spectrum, enabling the decoder to learn under varying degrees of spectral corruption and missing-fragment severity. Second, we introduce a mass-aware inductive bias into self-attention by using rotary embeddings to encode pairwise mass differences between spectral peaks. Together, these components align missing-fragment recovery with decoder behavior while explicitly incorporating the mass relationships that underlie peptide fragmentation. At inference time, GyroNovo retains a standard encoder-imputer-decoder architecture and requires neither additional inputs nor auxiliary search procedures. Experiments on NovoBench show gains of about 9 percentage points in peptide-level precision and 7 percentage points in amino-acid-level precision over the state-of-the-art baseline. Code: https://github.com/UBC-NLP/gyronovo. --- ### AutoResearch at Production Scale: Failure Modes and a Multi-Agent Framework Authors: Aparajith Chandran, Juwon Kim, Saurav Jha, Pablo Castells, Florian Hottier Published: 2026-09-24 Categories: cs.LG, cs.IR arXiv: https://arxiv.org/abs/2609.30541v1 Optimizing embedding systems for production recommendation pipelines demands systematic exploration that consumes disproportionate engineering effort at scale. We apply Andrej Karpathy's AutoResearch paradigm -- a large language model that iteratively edits a training script and retains modifications that improve a held-out scalar metric -- to automate this exploration. We report on twelve weeks of running this paradigm at production scale, where iterations consume hours of multi-GPU compute, evaluation involves competing criteria, and campaigns span weeks across many training jobs. Across two independently developed representation-learning systems for a book recommendation pipeline, we ran 220+ experiments and observed five recurring failure modes absent from the original setting: infrastructure fragility, agent memory decay, search-direction stagnation, iteration-cost asymmetry, and metric fixation. We contribute a three-principle scaffolding design -- prevent, persist, redirect -- that maps each failure mode to a structural remedy and whose instantiation scales with iteration cost. The framework produced a 1.82x Recall@6 lift and a 2.1x coherence lift over hand-tuned baselines, and the agent autonomously designed a text-only fallback that expanded catalog coverage by 5.8x. The two systems span nearly three orders of magnitude in per-iteration cost yet exhibit the same failure modes, suggesting these are structural properties of production-scale autonomous research rather than artifacts of either application. --- ### Don't CLAP: Are Music-Text Models Bag-of-Words? Authors: Yuan-Chiao Cheng, Alexander Lerch Published: 2026-09-24 Categories: cs.SD, cs.CL, eess.AS arXiv: https://arxiv.org/abs/2609.30540v1 Text-to-music systems are assessed on audio quality and on how faithfully the music follows its prompt, and the CLAP score, the cosine similarity between a music-text model's audio and text embeddings, is the standard objective metric of faithfulness. We ask how accurately that score reflects the text: when an attribute is linked to an instrument (e.g., distorted guitar), does the text embedding capture that binding? To find out, we introduce an attribute swap perturbation: the caption of a real recording is edited by exchanging exactly one property, timbre, lead versus accompaniment, or order of first appearance, between two instruments. We then test four contrastive music-text models and one large audio-language model on whether the audio scores higher against the original caption than against the perturbed one. No contrastive model distinguishes the two captions reliably. The audio-language model does better, but further experiments show that its advantage rests largely on audio-agnostic language priors. Our results thus provide compelling evidence that the CLAP score and related metrics do not capture fine-grained musical meaning or attribute bindings; their representation is closer to a bag-of-words that leaves them insensitive to meaning-changing perturbations of the caption. --- ### Learning to Replace MCMC in Split-Gibbs Diffusion Posterior Sampling via Deep Unfolding Authors: Yi Zhang, Rui Guo, Mengchu Xu, Zhaofeng Liu, Yonina C. Eldar Published: 2026-09-24 Categories: eess.SP, cs.LG, stat.ML arXiv: https://arxiv.org/abs/2609.30539v1 Split Gibbs sampling enables diffusion posterior inference for general nonlinear inverse problems by decoupling prior and likelihood computations, allowing a pretrained diffusion prior to be reused across measurement models. However, its likelihood update often relies on iterative MCMC, which can hinder parallelization, require algorithm-specific tuning, and incur substantial computational cost. In this work, we propose a learning-based framework to replace this MCMC step by reformulating both Gibbs updates as Gaussian denoising problems and implementing them through ODE diffusion. The prior step reuses a pretrained denoiser, while the likelihood denoiser exploits known likelihood structure through a lightweight deep-unfolded network. Experiments on nonlinear phase retrieval demonstrate the effectiveness of the proposed method as an alternative to MCMC-based split Gibbs at lower likelihood-update cost. --- ### Feeding BabyLMs Macaroni: Code-Switching Curricula Cause Cross-Lingual Convergence Authors: Dries Rooryck, Alex Cai, Yonatan Belinkov, David Alvarez-Melis, Kianté Brantley Published: 2026-09-24 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.30535v1 Children in multilingual communities often code-switch, using multiple languages in a single utterance. Can we induce cross-lingual alignment in language models by training on code-switched text? We pretrain small decoder-only transformers on two 100M-word multilingual corpora: a base corpus formed by mixing the English, Dutch, and Chinese BabyBabelLM datasets, and a corpus generated from it by inserting word- and sentence-level code-switching using an LLM. We find that training on code-switched data aligns the representations of parallel text, particularly across different scripts, and that this alignment persists through training on monolingual documents. Under a learning curriculum that progresses from word-level code-switching, to sentence-level code-switching, to monolingual documents, models trained on code-switched data outperform baselines trained without it on the BabyLM evaluation suite. Our work characterizes code-switching curriculum learning as an effective data augmentation method for multilingual pretraining. We release our code, data, and models at https://github.com/drooryck/multilingual-macaroni. --- ### Seeing Speech: Learning Visible Articulatory Dynamics for Speech-Driven 3D Facial Animation Authors: Hyung Kyu Kim, Byungchan Hwang, Hak Gu Kim Published: 2026-09-24 Categories: eess.AS, cs.GR, cs.LG arXiv: https://arxiv.org/abs/2609.30517v1 Recent progress in speech-driven 3D facial animation has improved vertex-level reconstruction quality, but speech-consistent visible articulation remains difficult. This is because speech production follows structured and constrained articulators' coordination and the mapping from acoustics to motion is inherently one-to-many. Motivated by the structured patterns of visible articulation, we propose a novel articulation-aware framework that models visible speech through directional articulatory motions and composes them into surface-consistent 3D facial motion. To represent visible articulation with three directional articulatory motions, spreading, opening, and protrusion, we propose a Speech--Articulatory Memory (SAM) that captures the correspondence between speech and these motions under phonetic context through retrieval and decoding based on a key-value memory structure. Then, a Topology-aware Articulatory Composition (TAC) integrates the predicted directional articulatory motions under mesh topology to produce surface-consistent 3D facial motion. Experiments on VOCASET and TFHP show that our method achieves state-of-the-art performance on standard reconstruction metrics and improves visible articulatory distance and velocity errors for lip articulation, while a user study confirms clear preference in lip sync and realism. --- ### Inquesto Score: A reliability Protocol For Voice Agents Authors: Massa Baali, Bhiksha Raj Published: 2026-09-24 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2609.30514v1 Voice agents are increasingly deployed in workflows where failed interactions can affect transactions, access, and other consequential outcomes, creating a need for reproducible and interpretable evaluation. We introduce Inquesto Score (IS), a protocol for measuring voice-agent reliability as the percentage of calls in a fixed, versioned evaluation population that achieve the caller's goal without a functional failure or worse. Rather than combining heterogeneous metrics, IS defines explicit failure events and severity levels and evaluates the deployed voice pipeline. Timing failures, including talk-over and delayed responses, are measured directly from audio, while semantic and state-dependent failures are evaluated using scenario predicates, tool traces, and a pinned open-model judge. Diagnostic views of behavior, acoustic robustness, identity handling, and speaker groups accompany the score without being combined into it. Inquesto Score v0.1 evaluates 30 scenarios, three acoustic conditions, four speaker groups, and 306 calls per agent across 13 configurations of a reference voice-agent system. Our evaluation shows that reliable measurement requires evidence beyond transcripts, explicit treatment of deployment conditions, and validation of the evaluators used to determine outcomes. We release the protocol, reference implementation, and evaluation records. ---