# Research Papers — 2026-07 · bundle page 38 of 222 > Full-text markdown of papers #1851–1900 of 11075 published in 2026-07, newest first. AI-readable mirrors of the full 11075-paper 2026-07 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-07/page/38 [← Previous bundle](https://aipod.dev/papers/archive/2026-07/page/37.md) [Next bundle →](https://aipod.dev/papers/archive/2026-07/page/39.md) ### Where Steering Signals Come From: Activation Source Selection in Activation Steering Authors: Jiaran Ye, Lingxu Ran, Zijun Yao, Chenpeng Wang, Yong Jiang, Lei Hou, Juanzi Li, Liangming Pan Published: 2026-07-28 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2607.25270v1 Activation steering controls language models by adding vectors or features to hidden states at inference time, but the upstream source of these steering signals is often treated as a secondary detail. We study this source choice as activation source selection: the combination of source context and activation readout policy used to collect the hidden states from which a steering signal is built. Holding the downstream intervention fixed, we show across three instruction-tuned models and four steering task families that changing only the source activations substantially changes steering success. We further find that effective steering is not explained simply by whether the desired behavior appears in the source text. Instead, strong signals come from execution-boundary states, where the model is about to produce or continue the target behavior. This pre-/post-realization distinction explains why answer-based sources sometimes work: their useful component aligns with execution-boundary directions rather than target appearance alone. Building on this view, we introduce tail subtraction, which removes shared prompt and continuation semantics from boundary states and yields cleaner, more stable steering signals. Overall, our results suggest that steering depends on representations of what the model is about to do, not merely on what has already appeared. --- ### Structure-aware Relative Policy Optimization for Ranking Authors: Yiteng Tu, Weihang Su, Zitao Su, Yiqun Liu, Min Zhang, Qingyao Ai Published: 2026-07-28 Categories: cs.IR, cs.AI arXiv: https://arxiv.org/abs/2607.25268v1 Ranking is a fundamental component of modern information access systems. Reinforcement learning (RL) provides a flexible framework for directly optimizing coarse-grained feedback and system-level objectives defined over the complete ranking list. However, existing RL-based ranking methods typically treat each sampled permutation as an atomic output and evaluate it primarily through a scalar reward, overlooking the structural relationships among different ranking lists. Consequently, permutations with similar rewards but substantially different permutation patterns may receive comparable optimization signals, potentially leading to inaccurate credit assignment and overly aggressive policy updates. To address this limitation, we propose SRPO, a \textbf{S}tructure-aware \textbf{R}elative \textbf{P}olicy \textbf{O}ptimization framework for listwise ranking. SRPO measures the discrepancy between sampled permutations using a top-weighted Kendall-tau distance and normalizes their pairwise reward differences by the corresponding distances. It quantifies the reward improvement per unit of ranking change, thereby emphasizing efficient local refinements, particularly those involving top-ranked positions. Experimental results across two ranking scenarios demonstrate that explicitly modeling permutation-level differences improves the effectiveness and stability of listwise ranking, with particularly favorable performance in limited-feedback and complex list-level optimization settings. --- ### FORGE: Frame Orthogonality in Relevance Geometry for Long-Form Video Understanding Authors: Ghazal Kaviani, Ghassan AlRegib Published: 2026-07-28 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2607.25266v2 Multimodal large language models (MLLMs) have enabled long-form video understanding at a scale that was not previously possible. However, the density of relevant content decreases sharply as video sequence length increases, and exposing the model to more irrelevant content measurably reduces its accuracy. In this paper, we address the problem of maximizing query-relevant information in a frame subset selected at inference time, without training. FORGE (Frame Orthogonality in Relevance Geometry) is a model-agnostic method that induces a query-conditioned geometry on a pretrained multimodal embedding space, unifying relevance and diversity into a single objective. In this space, frames that cover independent query-relevant directions are far apart, and selecting the subset of maximum information captures diverse query-relevant content within the budget. Experiments on Video-MME and LongVideoBench at budgets of 16, 32, and 64 frames show that FORGE improves the unified keyframe selection score by 11.0-15.3 points over the strongest training-free baseline and up to doubles keyframe recall (0.415 vs. 0.204 at K=64 on Video-MME). The gains extend to question answering, where accuracy improves in every evaluated setting across eight open-source MLLMs spanning 4B to 32B parameters, by up to 8.7 points over uniform sampling and 5.2 points over the strongest baseline. Our findings suggest that aligning the embedding space with the query's high-dimensional structure is a promising direction for inference-time video understanding. --- ### Laplace-PSN-IRT: Uncertainty Quantification for Neural Item Response Theory Models of LLM Benchmarks Authors: Juan Francisco, Mandujano Reyes Published: 2026-07-28 Categories: stat.AP, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2607.25257v1 Item Response Theory (IRT) has recently been proposed as a framework for evaluating large language model (LLM) benchmarks by separating a model's latent ability from the properties of individual benchmark items. Existing neural IRT approaches, including PSN-IRT, estimate these quantities using point estimates, limiting uncertainty quantification and downstream statistical inference. We introduce Laplace-PSN-IRT, a post-hoc last-layer Laplace approximation that augments a trained PSN-IRT model with approximate Bayesian posterior inference, recovering calibrated uncertainty over model ability and item difficulty without retraining. The resulting posterior enables credible intervals, probabilistic comparisons between models, and propagation of parameter uncertainty into Fisher-information-based item selection. We show that most pairwise comparisons among 12 models on a standard LLM benchmark leaderboard are not statistically distinguishable despite differing point-estimate ranks. We further show that point-estimate Fisher information can become nearly zero for many benchmark items because it is evaluated at a single reference ability, whereas posterior-expected Fisher information remains substantially more stable across the ability range. Finally, posterior-expected Fisher information more accurately recovers full-benchmark ability rankings from small benchmark subsets in most experimental settings while matching point-estimate performance for the smallest subsets. We validate the calibration of the approximate posterior using held-out predictive coverage and find that modeling item difficulty as random while treating item discrimination as fixed produces well-calibrated uncertainty in this architecture. --- ### The User Asks, Platforms Compete: How Agentic Recommendation Markets Take Shape Authors: Deyao Hong, Kehan Zheng, Qian Li, Jun Zhang, Jie Jiang, Hongning Wang Published: 2026-07-28 Categories: cs.AI, cs.IR arXiv: https://arxiv.org/abs/2607.25253v2 Online recommendation has traditionally taken place after a user enters a platform, which determines the candidate pool and the ranking shown to the user. LLM-based user agents enable a different recommendation process: a user specifies a need before choosing a platform, leaving platforms to compete for the user's attention, which we refer to as an agentic recommendation market. In our controlled LLM-based experiments across three product domains, we find this new setting of recommendation creates a tension between access and attention. Compared with traditional platform-centric recommendation, user-centric recommendation greatly expands the opportunity for relevant items to enter comparison; yet broader participation does not translate directly into effective exposure. Competition directly triggers platforms' strategic play: selectively positive explanations occupy 73--78% of first-ranked positions. When the user agent relates platforms' actions to subsequent user feedback, this share falls to 36--41%, while the chance of a user purchasing the relevant item increases. A user agent is therefore more than a ranker over a larger pool of candidates: its querying, ranking, and feedback mechanism governing who can compete, how scarce attention is allocated, and how earlier outcomes shape the evaluation of platforms directly affect user utility. Designing agentic recommendation therefore requires treating access, attention, and accountability as a joint mechanism design problem. --- ### CADENCE: A Cardiac Atom Dictionary for Interpretable Neural Concept Extraction from ECG Foundation Models Authors: Yixuan Duan, Arjun Naik, Sadeer Al-Kindi, Wei Qiu Published: 2026-07-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.25244v1 Foundation models for 12-lead electrocardiograms (ECGs) transfer well across clinical tasks, but the physiological knowledge encoded in their representations remains opaque. We present CADENCE, a framework that decomposes an ECG foundation model into a human-interpretable, queryable dictionary of physiological concepts. Using a BatchTopK sparse autoencoder, CADENCE factorizes Layer-6 embeddings from more than nine million ECG tokens into 8,192 sparse cardiac atoms. These atoms align better than individual dense embedding dimensions with clinical phenotypes and waveform morphology, recovering arrhythmias, conduction abnormalities, infarction and repolarization patterns, chamber and axis findings, and lead- and beat-phase-specific waveform primitives. At Layer 6, the best atoms achieve mean AUROCs of 0.88 for clinical phenotypes and 0.90 for morphology, versus 0.78 and 0.83 for the best dense dimensions. Sparse atom probes match or outperform dense probes for phenotype, morphology, and age prediction while attributing each prediction to a small set of interpretable atoms; phenotype AUROC improves from 0.93 to 0.95. Atom-space geometry recovers physiologically coherent relationships, and targeted atom ablation selectively changes frozen downstream outputs. An automated LLM pipeline generates and quantitatively validates atom descriptions by predicting held-out activations. On independent external ECG datasets, CADENCE recovers overlapping concepts and maintains consistent phenotype-prediction performance. CADENCE provides a scalable framework for discovering and auditing the physiological knowledge encoded by ECG foundation models. --- ### Medical world models in healthcare: foundations, applications, and challenges for trustworthy clinical translation Authors: Zhaoyan Chen, Zhongxiu Cong, Zhuanfeng Jin, Wanshu Fan, Dongsheng Zhou, Qi Ai, Haifan Gong, Congyu Liao, Xiaofeng Liu, Cong Wang Published: 2026-07-28 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.25242v2 Medical world models offer a framework for extending medical artificial intelligence beyond static prediction by representing evolving patient states and modelling how they change over time and in response to clinical interventions. This Review defines the conceptual boundaries, technical foundations, application domains, and evidence requirements of the field through a structured narrative synthesis with reproducible evidence mapping. We screened 1,455 unique records and assembled a corpus of 98 sources, including 14 studies that met a strict empirical definition of a medical world model. The field is organised around four capabilities: patient state representation, temporal dynamics modelling, intervention-conditioned simulation, and clinician-supervised planning. Evidence spans medical imaging, longitudinal electronic health records, treatment response modelling, physiological and multimodal state modelling, ultrasound and surgical interaction, and population and health-system simulation; clinical digital twins are treated as a cross-cutting integration framework. Current studies provide early evidence of technical feasibility for trajectory forecasting and comparison of candidate interventions, but most remain retrospective, task-specific, or preclinical. The evidence base is further limited by incomplete longitudinal intervention data, inconsistent action semantics, limited causal identifiability, long-horizon error accumulation, inadequate uncertainty estimation, and limited external validation. Clinical translation will therefore depend on precise intervention representations, robust causal and mechanistic grounding, calibrated trajectory-level uncertainty, safety-constrained planning, and prospective multicentre validation against clinically meaningful endpoints. --- ### Learning from the Unseen: Offline Reinforcement Learning with Hidden Actions Authors: Zeyu Bian, Ying Zhou, Yifan Cui Published: 2026-07-28 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2607.25241v1 Standard offline reinforcement learning (RL) algorithms typically assume that the actions in the dataset are observed without error. However, in many real-world applications, the true actions are unobserved and only noisy proxies are available, causing existing RL methods to yield biased and potentially misleading conclusions. We study off-policy evaluation in infinite-horizon discounted Markov decision processes with hidden actions. By leveraging the next-state variable as a natural proxy for the unobserved action, we establish identification of the policy value and propose an influence-function-based estimator called LURE (Learning from the Unseen: Robust Estimator). LURE is multiply robust, remaining consistent under several combinations of correctly specified nuisance components, and is asymptotically normal, enabling valid statistical inference. To our knowledge, this is the first work to address offline RL with hidden actions. We demonstrate LURE's effectiveness through simulations and a sepsis management application using the MIMIC-III database. --- ### CD-RMOT-Bench: Benchmarking the Cross-Domain Referring Multi-Object Tracking Authors: Xiangqun Zhang, Likai Wang, Zekun Qian, Ruize Han, Wei Feng Published: 2026-07-28 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.25239v1 Referring multi-object tracking (RMOT) extends tracking from category-driven perception to language-guided understanding by grounding object trajectories in natural-language expressions. Despite recent progress, existing RMOT studies are largely conducted under in-domain settings, leaving the robustness of language-conditioned tracking under inevitable visual domain shifts unexplored. In this paper, we study Cross-Domain Referring Multi-Object Tracking (CD-RMOT), a new and challenging problem that evaluates whether an RMOT model trained on a labeled source domain can reliably follow natural-language expressions in an unlabeled target domain with different visual conditions. To support systematic study, we construct CD-RMOT-Bench, a unified benchmark that combines real clear-domain referring tracking data, aligned digital-twin variants, and real adverse-domain videos. CD-RMOT-Bench enables both controlled weather/viewpoint shift analysis and realistic synthetic-real transfer evaluation under a shared RMOT protocol. Further, we provide a Query-Centric Adaptation (QCA) framework, designed to stabilize the query space that bridges visual trajectories and referring expressions. Extensive experiments reveal that domain shifts severely degrade RMOT performance, where the failure is not merely caused by object detection errors but more critically by unstable expression-conditioned temporal association and target selection. QCA establishes a strong baseline, while CD-RMOT-Bench opens a new direction for robust language-guided tracking across visual domains. --- ### VisualPatchWorld: Code World Models as Latent Structured Representations for Planning Authors: Jiaxin Bai, Jiaxuan Xiong Published: 2026-07-28 Categories: cs.CL, cs.RO arXiv: https://arxiv.org/abs/2607.25236v1 Different research lines use the term world model in different ways, yet they share a common aim: to capture how the world evolves under action in a form that supports perception, simulation, and planning. Two prominent realizations are neural predictors that learn dynamics in continuous vector spaces, and hand-built physics engines that expose explicit state and physical laws. Neural predictors scale from data but leave the form of the dynamics implicit; physics engines are inspectable and editable but difficult to construct at scale. We introduce VisualPatchWorld (VPW), which represents world dynamics as code. VPW first selects a qualitative dynamical form with short active probes, then fits that form's free parameters from recorded state-action traces by minimizing multi-step prediction error. The resulting programs can be rolled forward like a simulator, inspected in source form, and used inside model-predictive control; image-derived scene graphs can supply the live state at replan time. Across comparisons with prior code-based world models, VPW attains 69.0% mean planning success and exceeds the strongest code baseline by 23.5 points. The largest gains arise when choosing the correct qualitative dynamics is essential. Under the same planner, the induced models approach ground-truth engine success on navigation and grasp-rich control; a residual gap remains for contact-rich pushing, and checking a shortlist of promising plans in the engine closes most of that gap. These results establish a practical route toward automatically constructed code world models that are useful for planning. Code is available at https://github.com/HKBU-KnowComp/VisualPatchWorld/. --- ### WHTMix: Efficient Stereo Depth Estimation via Walsh-Hadamard Token Mixing Authors: Prathyush Sajith, Emadeldeen Hamdan, Ahmet Enis Cetin Published: 2026-07-28 Categories: cs.CV, eess.IV, eess.SP arXiv: https://arxiv.org/abs/2607.25234v2 Stereo depth estimation for driving, robotics and augmented reality must run at high resolution under tight latency budgets, yet in transformer-based matchers the global self-attention that aggregates scene context grows quadratically with the number of pixels and comes to dominate runtime. We show that the joint self-attention stage of a stereo transformer, whose role is to spread context across both views, can be replaced by a data-independent Walsh-Hadamard token mixer that mixes tokens globally in the transform domain at log-linear cost, while the data-dependent cross-attention that performs left-right correspondence is retained. On synthetic driving data the mixer matches the attention baseline in end-point error while reducing model compute by a factor of 2.46 and single-image inference latency by a factor of 2.65. A complexity analysis shows the benefit is governed by the ratio of sequence length to channel width, which explains why high-resolution stereo matching is a particularly favorable setting and why classification transformers are not; we confirm this token-to-channel scaling on non-stereo long-sequence benchmarks. Furthermore, we introduce a hybrid log-disparity loss function designed to up-weight small-disparity pixels corresponding to long-range objects. This approach reduces the error on distant objects without incurring any additional computational overhead. --- ### Beyond Single-Episode Optimization: Sliding-Window Aware Generative Auto-Bidding for Long-Term Advertising Effectiveness Authors: Binglin Wu, Chuan Yue, Yingyi Zhang, Xianneng Li, Ruyue Deng, Weiru Zhang, Xiaoyi Zeng Published: 2026-07-28 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.25233v1 Auto-bidding systems optimize bids to maximize value under efficiency constraints such as Cost-Per-Action (CPA). Existing methods treat each day as an independent episode. However, many advertisers produce value so sparsely that per-day efficiency ratios become statistically unreliable, undermining advertiser retention. Platforms therefore evaluate window-level efficiency over sliding windows of $W{=}7$ days, ensuring fair evaluation and long-term advertising effectiveness. This creates cross-episode coupling: each day's bidding decisions affect up to $W$ overlapping windows, so setting daily targets requires anticipating future market conditions. We propose SWAG-Bid (Sliding-Window Aware Generative Auto-Bidding), a hierarchical framework decomposing this challenge into episode-level planning and step-level execution. The planner uses a Masked Trajectory Model to forecast markets and generate candidate plans, scored across all overlapping windows by Multi-Window Model Predictive Control Sampling (MWMS) with exponential confidence decay. The controller adjusts reliance on this guidance through a state-adaptive gate, Per-Step Gated Adaptive Layer Normalization (PSG-AdaLN), complemented by Return-to-Go and Cost-to-Go channels carrying budget and constraint information. Experiments on AuctionNet-Sparse and online A/B tests on AliExpress show that SWAG-Bid achieves competitive constraint satisfaction and value acquisition under sliding-window evaluation. --- ### Neurai-VN Benchmark: Standardized Machine Learning Models for Multimodal Digital Phenotyping in Mental Health Classification Authors: Quoc-Cuong Pham, Hoang-Thuy-Duong Vu, Thi-Thanh-Huong Ha, Huy-Hieu Pham Published: 2026-07-28 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.25232v3 Digital phenotyping (DP) using smartphones and wearable devices has emerged as a promising approach for assessing mental health, particularly depression and anxiety. However, progress remains difficult to evaluate because of heterogeneity across datasets and inconsistencies in preprocessing pipelines. In this work, we introduce a reproducible machine learning benchmark using the Neurai-VN dataset, a multimodal digital phenotyping dataset collected from 100 Vietnamese adults over two weeks. We define four binary classification tasks evaluated using standardized subject-wise cross-validation. Representative linear, tree-based, and neural baseline models are evaluated systematically across predefined feature-group configurations. Mean subject-level F1 scores across five cross-validation folds reached 0.71 for Healthy Control vs. Depression and Healthy Control vs. Clinical, while Healthy Control vs. Anxiety and Depression vs. Anxiety achieved 0.69 and 0.56, respectively. These baseline results provide reproducible baselines for future research on multimodal DP for mental health classification tasks. The code to reproduce the benchmark is available at https://github.com/neurai-vn/Neurai-VN-benchmark. --- ### Neurai-VN Benchmark: Standardized Machine Learning Models for Multimodal Digital Phenotyping in Mental Health Classification Authors: Quoc-Cuong Pham, Hoang-Thuy-Duong Vu, Thi-Thanh-Huong Ha, Huy-Hieu Pham Published: 2026-07-28 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.25232v1 Digital phenotyping (DP) using smartphones and wearable devices has shown considerable potential for mental health monitoring. However, progress remains difficult to evaluate due to heterogeneous datasets, inconsistent preprocessing pipelines. In this study, we present a reproducible benchmark built upon the Neurai-VN dataset, a high-resolution, multimodal dataset comprising passive sensing and active assessment from wearable and smartphone devices, collected from 100 Vietnamese adults over two weeks. The benchmark defines four clinically relevant binary classification tasks evaluated using standardized subject-wise cross-validation. Representative linear, tree-based, and neural baseline models are evaluated across predefined feature configurations. Mean subject-level F1 scores across five cross-validation folds reached 0.71 for Healthy Control vs. Depression and Healthy Control vs. Clinical, while Healthy Control vs. Anxiety and Depression vs. Anxiety achieved 0.69 and 0.56, respectively. These benchmark results provide reproducible baselines for future research on multimodal DP for mental health classification tasks. --- ### Interpretable Column Annotation with LLM-Symbolized Decision Process Materialization Authors: Mengqi Wang, Jianwei Wang, Qing Liu, Xiwei Xu, Zhenchang Xing, Michael Bain, Liming Zhu, Wenjie Zhang Published: 2026-07-28 Categories: cs.CL arXiv: https://arxiv.org/abs/2607.25228v1 Column annotation (CA), including column type annotation (CTA) and column property annotation (CPA), aims to identify the meanings of table columns and the semantic relationships among them. Recent CA methods usually use various neural models to learn column representations and directly map them to label categories, thereby (1) sacrificing model interpretability and adaptivity, and (2) overlooking rich label semantics and ultimately limiting accuracy. To address these limitations, we propose SymCA, an LLM-empowered interpretable CA framework that materializes column annotation as a global-to-local symbolic decision process. SymCA consists of two components: (1) global skeleton induction, which constructs a semantic skeleton over the label space, and (2) local substrate evolution, which evolves predictive substrates within the skeleton. Specifically, to exploit label semantics while preserving an interpretable decision process, the global skeleton induction module leverages LLMs to generate candidate hypernym-inspired tree-structured semantic skeletons and employs a Minimum Bayes Risk (MBR)-based consensus strategy to select a robust skeleton against generation variance. Since different internal nodes require different evidence to distinguish among their child nodes, the local substrate evolution module materializes each internal node as an executable and evolvable predictive substrate. Over multiple evolution rounds, each substrate trains an interpretable random forest classifier with the current operator set, leverages the LLM to propose node-specific operator modifications, and uses an exploration-exploitation strategy to prioritize promising substrates. Extensive experiments demonstrate that SymCA is accurate, robust, and interpretable, outperforming the strongest baselines by an average of 6.42% in Micro-F1 and 11.03% in Macro-F1. --- ### Decision-Level Hijacking: Injecting Cognitive Bias into Large Language Models via Bit-Flip Attacks Authors: Yu Yan, Jiahao Chen, Siqi Lu, Yongjuan Wang, Ziming Zhao, Zhaoxuan Li, Tianyu Du, Qingjun Yuan, Shouling Ji Published: 2026-07-28 Categories: cs.CR, cs.LG arXiv: https://arxiv.org/abs/2607.25227v1 Large Language Models (LLMs) have been widely applied in high-stakes decision-making scenarios such as corporate strategy, and users are increasingly relying on their outputs. However, the deep integration of open-source model sharing ecosystems with LLM-powered critical decision-making applications also introduces critical risks: if an attacker can manipulate the model's cognitive stance, they can indirectly influence the judgments and actions of downstream decision-makers. This paper defines such threats as decision-level hijacking. Existing attacks fail to achieve targeted cognitive manipulation without triggering prohibited content or degrading model functionality. To fill this gap, this paper reveals that Bit-Flip Attacks (BFAs) can serve as an attack vector for inducing decision-level hijacking, requiring no real-time interaction or control over the training process, and only a minimal number of weight bits need to be flipped after deployment to achieve stealthy, low-cost, and persistent cognitive manipulation. Therefore, we propose CogBias, a cognitive bias injection framework for LLMs. CogBias converts subjective preferences into optimization signals via a differentiable sentiment evaluator, uses a multi-objective loss to jointly constrain multiple dimensions, and constructs BitScout to locate critical bits, achieving targeted cognitive intervention under an ultra-sparse flip budget. Experiments on Llama-3.2-3B, Mistral-7B, and Qwen2.5-14B, as well as on the commercial recommendation and controversial factual topic scenarios, demonstrate that flipping only a small number of bits stably induces significant stance shifts on target topics, while the impact on non-target tasks and overall output distribution is limited. This work demonstrates that minute perturbations to low-level weight data suffice to undermine the high-level value alignment of LLMs. --- ### SecDrift: Measuring Sector-Conditioned Security Drift in AI-Generated Code Authors: Narayanaswami Natraj Bharadwaj, Dhivya Chandramouleeswaran Published: 2026-07-28 Categories: cs.CR, cs.LG, cs.SE arXiv: https://arxiv.org/abs/2607.25225v1 LLMs are increasingly used for code generation in critical infrastructure, yet the security effect of domain-specific prompting is understudied. We present SecDrift, a benchmark measuring sector-conditioned security drift: the change in static-analysis vulnerability rates when prompts are conditioned on industry contexts versus neutral baselines. We evaluate 7 LLMs (6 producing analyzable code) across 8 CISA critical infrastructure sectors and 9 CWE categories with 5 replicates (5,355 evaluations), using a 5-dimension transformation with a matched-baseline condition that holds the task fixed while substituting only domain terminology. Industry prompts naively appear more secure (14.0% vs. 11.4%, -2.7pp), but the gap is not statistically significant (Fisher's exact p = 0.24, Cohen's h = -0.08) and is a composition artifact of two CWE categories: excluding CWE-502 and CWE-22 eliminates and slightly reverses it (+0.4pp, p = 1.00). A mixed-effects logistic regression confirms sector identity is not a moderator and localizes the only detectable condition effect to those two vulnerability types. 0 of 8 sectors show drift distinguishable from baseline, corrected or uncorrected (|h| < 0.15). A placebo on two non-CISA sectors (e-commerce, online education) reproduces the CISA industry rate almost exactly (10.5% vs. 11.4%, p = 0.63): the small pooled pattern reflects generic industry-framing specificity, not critical-infrastructure identity. In contrast, model selection has a large and consistent effect: among full-output models vulnerability rates range from 11.6% to 16.1%, and these differences persist across conditions. Model choice, not prompt framing, is the more reliable security lever. We release the framework, prompts, generated code, findings, human-validation verdicts, and analysis scripts. --- ### Everyone is unique: Towards Behaviorally Heterogeneous Negotiation Dialogue Systems for Debt Collection Authors: Yuhang Yang, Kai Tang, Chao Ye, Haobo Wang, Qiqi Luo, Jinguang Zheng, Zhixin Zhang Published: 2026-07-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.25218v1 Debt collection is a critical negotiation task in the financial industry, with strong practical relevance and exceptional academic value as a behaviorally rich, high-stakes testbed for human-centered dialogue systems. While large language models (LLMs) have shown promise in dialogue and negotiation, effectively evaluating their performance in this complex scenarios remains a major challenge: existing benchmarks uniformly assume users to be static, rational agents with fixed preferences, failing to capture the rich behavioral heterogeneity inherent in real-world debt collection. To bridge this gap, we propose DebtBench, the first public persona-enriched debt collection benchmark, that highlights behavioral heterogeneity in negotiation. Moreover, we develop DebtGPT, a debt collection agent trained to jointly optimize financial recovery and interaction experience. Our experimental results, using 16 state-of-the-art LLMs, find that most existing models struggle in this complex but realistic scenarios, whereas DebtGPT outperforms all open-source baselines and achieves performance on par with GPT-4o. The code and data are available at https://github.com/YYuHhhh/DebtNegotiation. --- ### TopoGR: Revealing and Preserving Latent Structure of Semantic ID in Generative Recommendation Authors: Ziyu Zheng, Zhengshun Du, Yaming Yang, Bin Tong, Guan Wang, Meng Yan, Ziyu Guan, Wei Zhao Published: 2026-07-28 Categories: cs.IR, cs.AI arXiv: https://arxiv.org/abs/2607.25216v1 Semantic ID-based generative recommendation tokenizes each item into a sequence of discrete semantic IDs and predicts the next item by generating semantic IDs. However, existing methods typically regard SIDs as independent discrete symbols, while often overlooking the topology of the learned semantic ID space. We identify a structural mismatch between tokenization and generation: the tokenizer learns a structured code space with semantic neighborhood relations, whereas the generator consumes semantic ID tokens as independent categorical symbols. Consequently, item relatedness is reduced to exact semantic ID overlap, making it difficult to identify semantically similar items whose semantic IDs do not overlap. To address this issue, we propose TopoGR, a topology-preserving generative recommendation framework based on Bit-decomposable Semantic ID(Binary SID). Each Binary SID is learned in a bit-decomposable form and can be deterministically converted to a standard integer SID, while exposing an explicit Hamming geometry. TopoGR exploits this topology at three stages: binary SID features preserve Hamming proximity at the input layer; Hamming soft targets inject topology-aware supervision; and Hamming-consistent reranking aligns candidate items with the predicted binary prototype during inference. We further verify that the Hamming topology can capture item relatedness beyond exact SID matching. Experiments on four benchmark datasets show that TopoGR consistently outperforms existing state-of-the-art baselines in recommendation performance. --- ### Leveraging Semantic Maps for City-Scale Cross-View Localization Authors: Ethan Fahnestock, Erick Fuentes, Philip R Osteen, Nicholas Roy Published: 2026-07-28 Categories: cs.RO, cs.CV arXiv: https://arxiv.org/abs/2607.25215v1 We want robots to localize in previously untraversed environments against commonly available prior data. Rich semantic data available from OpenStreetMap can be useful in this task. However, existing methods either ignore this semantic information, directly matching panoramas and overhead imagery, or dramatically compress the semantic information, working with a small set of fixed classes. To leverage this rich semantic information, two challenges need to be overcome. First, useful semantic information needs to be extracted from the robot's egocentric observations. Second, the observed information must be quickly associated with the large prior semantic map (e.g., up to 628 km^2). We show that VLMs are effective at both extracting relevant landmarks from panoramas, and identifying feasible correspondences between these landmarks and prior overhead landmarks. However, using VLMs to propose all correspondences scales poorly as the number of mapped landmarks increases. Instead, we propose distilling a lightweight matcher from a VLM which computes correspondences for all entities in a map. We use this output to form an observation likelihood which is fused over time with a Bayes filter to create a time series of pose estimates. To support further investigation into generalizable cross-view methods that leverage semantic information, we release a dataset of extracted semantics and evaluation trajectories spanning eleven environments, including panoramas we collected in a snowstorm and at night in Boston. We demonstrate our method, trained on a single city's fair-weather data, generalizes across location, lighting, weather, and other challenges. Code and datasets are available at https://efahnestock.github.io/loci/. --- ### ObliCity: A Benchmark and Baseline for Roof-to-Ground Projection Displacement Correction Authors: Kai Li, Yupeng Deng, Ligao Deng, Zhihao Xi, Chenhao Wang, Jierui Zhang, Yingrui Ji, Yu Meng, Xiangyu Zhao Published: 2026-07-28 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.25210v1 Oblique-view urban remote sensing imagery inevitably exhibits geometric projection displacements between building roofs and footprints, leading to significant distortions in spatial structure. Existing approaches either ignore these deformations or handle them implicitly within segmentation-based frameworks, where progress is dominated by general segmentation advances rather than improvements in geometric correction. In this work, we explicitly define roof-to-footprint offset vector (RFOV) extraction as an independent learning task that decouples geometric alignment from semantic segmentation. To support this task, we introduce the Oblique City dataset (ObliCity), the first large-scale benchmark that integrates high-resolution UAV imagery and globally distributed satellite data, covering diverse city morphologies and camera perspectives. Methodologically, we reformulate DragOSM into DragRoof, an ODE-based framework inspired by human annotation behavior. By simulating the continuous process of dragging roofs toward their footprints, DragRoof learns deterministic, geometry-consistent offset fields and adaptively determines convergence through an end token. Extensive experiments on ObliCity demonstrate that DragRoof achieves state-of-the-art RFOV extraction performance, requiring fewer inference steps while delivering superior directional and length accuracy. Our dataset and model establish a principled foundation for studying projection displacement correction in oblique remote sensing imagery. The source code and dataset will be avaliable at https://github.com/likaiucas/DragRoof. --- ### VaLiDRec: Variable-Length LLM-Aligned Semantic IDs for Generative Recommendation Authors: Shutong Qiao, Wei Yuan, Tong Chen, Hao Wang, Quoc Viet Hung Nguyen, Hongzhi Yin Published: 2026-07-28 Categories: cs.IR, cs.AI arXiv: https://arxiv.org/abs/2607.25209v1 Generative recommendation commonly represents items using fixed-length semantic identifiers (SIDs) constructed through clustering and quantization. However, these artificial codes may overcompress item semantics, remain misaligned with pretrained LLM vocabularies, and require costly autoregressive decoding. In light of this, we propose VaLiDRec, a generative recommendation framework based on variable-length, LLM-aligned semantic identifiers. VaLiDRec constructs SIDs directly from informative native LLM vocabulary tokens via token importance estimation, semantic-quality-aware pruning, and collision-aware refinement, allowing identifier lengths to adapt to item semantic complexity. To model user preferences, VaLiDRec incorporates graph-aware soft prompts and reformulates recommendation as token-set prediction with token-level item scoring, eliminating autoregressive SID generation and beam search. Experiments on four real-world datasets show that VaLiDRec consistently outperforms strong sequential and generative recommendation baselines across all evaluation metrics. It further achieves superior zero-shot item cold-start performance and 87.49$\times$ faster inference than LC-Rec. These results demonstrate that LLM-native variable-length semantic identifiers provide a more expressive and efficient paradigm for generative recommendation. --- ### A Unified Algorithmic Framework for Hybrid Reinforcement Learning in Tabular MDPs with Shifted Transition Dynamics Authors: Zheshun Wu, Renjie Zheng, Jinhang Zuo, Zenglin Xu, Fang Kong Published: 2026-07-28 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.25207v1 This paper investigates a hybrid reinforcement learning setting in tabular Markov Decision Processes (MDPs), where an agent aims to learn an optimal policy by combining online interactions with a target environment and offline data from a source environment. A central challenge is that offline data may be collected from outdated environments with shifted transition dynamics, making naive integration of historical data ineffective. To address this, we propose a unified algorithmic framework featuring two algorithms: MIN-UCB-VI for regret minimization and MAX-LCB-VI for best policy identification. Both algorithms leverage fine-grained bias information to more effectively exploit offline data under general transition shifts. We provide theoretical guarantees for our framework, including both instance-dependent and independent upper bounds on regret and sub-optimality gap. Furthermore, we establish matching lower bounds to demonstrate the optimality of our approach and validate our theoretical findings through extensive experiments. --- ### A Cross-lingual Comparison of Human and Classification Model Entrainment Behavior in Code-switched Speech Settings Authors: Debasmita Bhattacharya, Siying Ding, Alayna Nguyen, Julia Hirschberg Published: 2026-07-28 Categories: cs.CL arXiv: https://arxiv.org/abs/2607.25202v1 Conversational entrainment is well-studied in monolingual and written contexts, but remains underexplored in spoken code-switching (CSW). We present a novel cross-lingual analysis of entrainment in Mandarin-English, Hindi-English, and Spanish-English dialogue and show that, while lexical entrainment generalizes across language pairs, entrainment over acoustic-prosodic and CSW style aspects exhibits context-specific variation. We build on these findings by asking whether classification models capture these human behavioral patterns. Applying feature importance and ablation analyses, we find that classical and Transformer-based classifiers detect entrainment reasonably well but consistently prioritize features other than those most salient to human entraining behavior. Our approach introduces a human-grounded framework for evaluating model decision-making in multilingual stylistic contexts, and suggests future challenges for developing conversational agents capable of producing naturalistic code-switched speech. --- ### Algorithmic Separation between Constant-Depth and Logarithmic-Depth Neural Networks Authors: Yunwei Ren, Zihao Wang, Jason D. Lee Published: 2026-07-28 Categories: cs.LG, stat.ML arXiv: https://arxiv.org/abs/2607.25200v1 Despite the empirical advantages of deep networks over shallow ones, theoretical depth separations largely concern approximation power, while algorithmic results are mostly limited to comparisons between two- and three-layer networks. In this work, we prove the first algorithmic separation between constant-depth and logarithmic-depth networks. Specifically, we identify a class of Boolean functions with hierarchically structured Fourier spectra that logarithmic-depth networks can learn efficiently using layerwise coordinate descent by reconstructing the spectra hierarchically and adaptively. We also exhibit a subclass for which every constant-depth, polynomial-width network with sufficiently regular activations and controlled spectral norms must incur constant $L^2$ approximation error under the uniform distribution over the hypercube. --- ### RIDGE: An Autonomous Framework for Validation and Method Discovery in LLM-Generated Option Pricing Authors: Liexin Cheng, Xue Cheng, Shuaiqiang Liu, Cornelis W. Oosterlee Published: 2026-07-28 Categories: q-fin.CP, cs.AI arXiv: https://arxiv.org/abs/2607.25199v2 Automated code generation is becoming an important tool in quantitative finance, where large language models can generate option pricing implementations directly from mathematical model specifications. Validating such implementations, however, requires considerably more than conventional software testing: numerical pricing methods must remain mathematically consistent, numerically stable, and reliable across a wide range of model parameters. We introduce RIDGE, an autonomous validation framework in which generated pricing implementations are subjected to structured no-arbitrage tests, stress tests, benchmark comparisons, and consistency checks. Validation evidence is interpreted diagnostically, while the resulting knowledge is accumulated in a repository and reused across models and successive validation iterations. This enables systematic refinement of both the pricing implementation and the validation methodology. The framework is applied to five stochastic volatility models. Across these studies, all detected implementation defects are removed and, in two cases, the validation process reveals methodological limitations and motivates the development of alternative numerical methods. The supplementary material is available in the GitHub repository: https://github.com/ShQiangLiu/ridge. --- ### LGFNet: A CTC-Guided Local-Global Fusion Framework for Single-Channel Sleep Staging Authors: Chongjian Wang, Zhenghang Hou, Junjie Gao, Xiaofang Zhong, Shiyuan Han, Tong Zhang Published: 2026-07-28 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.25197v2 Sleep staging remains challenging due to long-range temporal dependencies, ambiguous stage transitions-particularly in N1-and substantial distribution shifts across subjects, sampling rates, and EEG montages. These difficulties are further amplified in single-channel, low-latency scenarios required by wearable and real-world applications. To address these issues, we propose LGFNet, a CTC-guided sequence-to-sequence framework for robust sleep staging. LGFNet introduces a Local-Global Fusion encoder that jointly models fine-grained temporal dynamics and long-range sleep structure, overcoming the limitations of conventional serial hybrid architectures. A CTC-Attention joint training paradigm is adopted to unify temporal alignment with context-dependent modeling, enabling more accurate recognition of stage boundaries and transitions. Furthermore, a three-stage decoding strategy is devised, leveraging CTC-guided decoding and Viterbi-based smoothing to reduce error accumulation and enforce physiological consistency. Extensive cross-dataset evaluations on five public benchmarks demonstrate that LGFNet consistently outperforms state-of-the-art single-channel methods. In particular, on Sleep-EDF-78, LGFNet surpasses DMIN by +1.27% accuracy, +1.74% macro-F1, and +1.93% kappa, with pronounced gains on N1 and transition segments, highlighting its robustness and strong generalization across diverse sampling rates, montages, and recording environments. --- ### LGFNet: A CTC-Guided Local-Global Fusion Framework for Single-Channel Sleep Staging Authors: Chongjian Wang, Zhenghang Hou, Junjie Gao, Xiaofang Zhong, Shiyuan Han, Tong Zhang Published: 2026-07-28 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.25197v1 Sleep staging remains challenging due to long-range temporal dependencies, ambiguous stage transitions-particularly in N1-and substantial distribution shifts across subjects, sampling rates, and EEG montages. These difficulties are further amplified in single-channel, low-latency scenarios required by wearable and real-world applications. To address these issues, we propose LGFNet, a CTC-guided sequence-to-sequence framework for robust sleep staging. LGFNet introduces a Local-Global Fusion encoder that jointly models fine-grained temporal dynamics and long-range sleep structure, overcoming the limitations of conventional serial hybrid architectures. A CTC-Attention joint training paradigm is adopted to unify temporal alignment with context-dependent modeling, enabling more accurate recognition of stage boundaries and transitions. Furthermore, a three-stage decoding strategy is devised, leveraging CTC-guided decoding and Viterbi-based smoothing to reduce error accumulation and enforce physiological consistency. Extensive cross-dataset evaluations on five public benchmarks demonstrate that LGFNet consistently outperforms state-of-the-art single-channel methods. In particular, on Sleep-EDF-78, LGFNet surpasses DMIN by +1.27% accuracy, +1.74% macro-F1, and +1.93% kappa, with pronounced gains on N1 and transition segments, highlighting its robustness and strong generalization across diverse sampling rates, montages, and recording environments. --- ### Rethinking CD: A Reproducibility Study and Extension on the Ineffectiveness of Contrastive Decoding at Mitigating Object Hallucinations in MLLMs Authors: Arnav Bendre, Guneesh Gupta, Kavish Grover, Chayan Aggarwal, Shreyansh Modi Published: 2026-07-28 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.25196v1 Contrastive decoding (CD) has been proposed as a training-free strategy for mitigating object hallucinations in multimodal large language models (MLLMs), with reported gains on benchmarks such as POPE. However, recent work has questioned whether these gains reflect genuine improvements in visual grounding. In this study, we reproduce and extend the findings of "The Mirage of Performance Gains: Why Contrastive Decoding Fails to Mitigate Object Hallucinations in MLLMs." Specifically, we test the claim that CD induces a unidirectional output distribution shift in discriminative datasets and examine its generalizability across datasets. We also verify that the adaptive plausibility constraint (APC) reduces sampling to greedy search on both discriminative and generative benchmarks. Beyond reproduction, we rigorously study the effects of CD across generative and discriminative datasets. We conduct several experiments that provide additional insights: we analyze the logit distributions induced by different CD strategies on generative datasets, propose a proxy method and compare its performance against CD techniques, and investigate how hallucination signals propagate through each layer of the expert and amateur models. Experimental results across MME, POPE, and CHAIR using LLaVA and Qwen validate the original claims and show that the apparent improvements from CD are often spurious and do not consistently translate into stronger visual grounding for reducing hallucinations. These findings challenge the effectiveness of current contrastive decoding strategies and motivate the development of more reliable approaches for mitigating hallucinations in MLLMs. --- ### Lloyd's $K$-Means Clustering Algorithm Is Frank-Wolfe in Disguise Authors: Michael Pokojovy, J. Marcus Jobe, Simon Lacoste-Julien Published: 2026-07-28 Categories: stat.ML, cs.LG, stat.CO arXiv: https://arxiv.org/abs/2607.25190v1 Lloyd's $K$-means algorithm, also known as naïve $K$-means, is a widely used ad hoc optimization heuristic, designed to minimize the sum of squared errors (SSE) across all $K$-partitions of a dataset via iterative cluster refinement. In this work, we establish a novel connection between Lloyd's algorithm and the Frank-Wolfe (FW) algorithm, a prominent first-order method for projection-free optimization. We demonstrate that Lloyd's algorithm is a special case of FW. Leveraging recent advances in FW methods for concave objectives, we derive a non-asymptotic $\mathcal{O}(1/t)$ convergence rate to a local minimum of the SSE objective. To account for empty clusters, an outcome possible under Lloyd's greedy assignment, we develop an FW variant for semismooth objectives while retaining the same convergence rate that is solely controlled by the initial SSE value. We illustrate our findings with a simulation study for spherical Gaussian mixtures and a real-world image segmentation dataset. --- ### CardioBench: A Real-World Data Benchmark for Evaluating Large Language Models in Clinically Authentic Cardiovascular Care Scenarios Authors: Xiao Li, Mouxiao Bian, Zhaodi Wu, Sijie Ren, Juechen Chen, Lu Lu, Jingru Ding, Yun Zhong, Jie Xu, Yixiu Liang, Junbo Ge Published: 2026-07-28 Categories: cs.CL arXiv: https://arxiv.org/abs/2607.25186v2 Background: Most medical large language model (LLM) benchmarks focus on examination knowledge or isolated tasks and may not reflect the longitudinal, multimodal, and safety-critical workflow of cardiovascular care. Objective: To develop CardioBench, a real-world benchmark spanning the cardiovascular care continuum, and assess LLM performance across clinical dimensions and specialist tasks. Methods: CardioBench includes 2,263 items from 13 task-specific datasets derived from de-identified cardiovascular records and examination data. Sixteen cardiology physicians conducted annotation and reference construction, followed by cross-review from two senior cardiologists. Seven LLMs generated 15,841 outputs under standardized zero-shot settings. Open-ended tasks were evaluated using key-point coverage and holistic clinical quality, while CardioEthics was scored by accuracy. Results: GPT-5.4 achieved the highest macro-average (62.55) and item-weighted mean (62.19), followed by Gemini 3.1 Pro (59.95) and Qwen 3.6 27B (59.72). GPT-5.4 ranked first in all three dimensions. CardioAuxReport performed best (86.38), whereas CardioECGRead (17.25) and CardioEthics (17.34) were lowest. The largest gaps between holistic clinical quality and key-point coverage occurred in CardioComm (52.71), CardioEmergRescue (52.05), and CardioTreatPlan (48.80). Conclusions: To our knowledge, CardioBench is the largest real-world, multi-task benchmark for LLM evaluation across the cardiovascular care continuum and offers the broadest coverage of clinically authentic cardiology scenarios reported to date. It provides a rigorous framework for identifying model strengths, clinically important omissions, and priorities for future development. --- ### A scaling law of contextual persistence in human language Authors: Elan Barenholtz Published: 2026-07-28 Categories: cs.CL arXiv: https://arxiv.org/abs/2607.25184v1 Human language exhibits lawful structure at the level of words (frequency, vocabulary growth) and word pairs (co-occurrence across distance). Here we show that the arrangement of words in sequence -- a central determinant of meaning -- obeys a comparable law. Using large language models as probabilistic probes, we measured the reduction in target perplexity conferred by prior context at distance d beyond that of the same words scrambled; this difference, the contextual persistence function P(d), isolates the influence of arrangement. Across ten corpora spanning six language families and written and spoken modalities, P(d) decayed approximately as 1/d ($P(d) \propto d^{-α}$, mean $α= 1.04$; median $r^2 = 0.96$). The effect vanished in scrambled and synthetic controls, replicated across independent probes, and did not appear in genomic or protein sequences under domain-native models. An exponent near 1 distributes contextual influence approximately uniformly across logarithmic timescales. The results establish a scaling law of contextual persistence in human language. --- ### TabRank: Chain-of-Thought Distillation for Table Re-Rankers Authors: Adarsh Singh, Kushal Raj Bhandari, Jianxi Gao, Soham Dan, Vivek Gupta Published: 2026-07-28 Categories: cs.CL, cs.AI, cs.IR arXiv: https://arxiv.org/abs/2607.25182v1 The ability to retrieve relevant tables for answering questions is a key task for structured information retrieval. Multi-stage retrieval systems rely heavily on rerankers to refine candidate lists produced by efficient first-stage retrievers. As a result, neural rerankers and LLM-based reranking methods have become increasingly important due to their superior capacity for semantic understanding and reasoning compared to conventional sparse or dense retrieval models. Recently, Large Reasoning Models (LRMs) equipped with explicit chain-of-thought (CoT) reasoning have shown strong improvements in ranking quality in unstructured passage retrieval. In this work, we present TabRank, a framework for training reasoning rerankers for Tabular Retrieval. We first present a comprehensive dataset of 6728 reasoning traces for tabular reranking on the Natural Questions Tables dataset. We then explore two variants of training a compact reasoning model on these reasoning traces: explicit CoT distillation and conditioning the student reranker on the teacher's reasoning trace within the prompt. We stress-test TabRank on several out-of-distribution generalization settings on diverse domains and multi-table scenarios. Our approach significantly improves performance across a variety of table retrieval datasets, increasing Acc@10 by 30.5% on HybridQA, 15.2% on SQA, 52.9% on TabFact, and 13.1% on TATQA subsets of the Multi-Table QA Benchmark compared to the base model. Notably, TabRank generalizes effectively to multi-table reasoning. Our code, data and models are available at https://github.com/AdarshSingh7647/TabRanker --- ### CondPSE: A Polynomial-Filtered Structural Encoder with Conditional Modulation for Graphs Authors: Woohyun Lee, Hogun Park Published: 2026-07-28 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2607.25169v1 Message-passing graph neural networks are bounded by the 1-WL test and can miss topological structure that distinguishes non-isomorphic graphs. Positional and structural encodings (PSE) inject such topology-derived signals, and learned PSE encoders such as GPSE pretrain a single encoder to produce these signals from random node probes, which can then be frozen and reused as inputs across downstream graph models. We present CondPSE, a learned PSE encoder that applies a learnable polynomial graph filter bank to standard Gaussian node probes and refines the resulting structural-response branches through FiLM-style modulation conditioned on cross-filter, local message-passing, and graph-level signals. CondPSE is pretrained to reconstruct node-level positional/structural targets and graph-level invariants, and is then frozen for use as a downstream input encoding. On synthetic structural-discrimination benchmarks, CondPSE separates graph structures that 1-WL-bounded message passing cannot: it raises CSL accuracy from 42.9% to 97.3% and EXP accuracy from 68.3% to 99.9% relative to GPSE, and ablations show that the polynomial filter bank accounts for most of this gain. On real molecular property prediction, the picture is more limited. With a hybrid local-message-passing/global-attention backbone, CondPSE performs comparably to GPSE without surpassing it, and a ZINC backbone sweep shows no consistent ordering between the two encoders. We report these results and discuss why strong synthetic structural discrimination does not, on its own, yield a downstream advantage for frozen learned PSE encoders, including the role of downstream integration and possible mismatch between structural pretraining targets and molecular property labels. --- ### Individual-level interventions against sycophantic AI reduce its appeal but not its persuasiveness Authors: Meryl Ye, Robert Kraut, Steve Rathje Published: 2026-07-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.25166v3 AI chatbots can be "sycophantic," or overly agreeable and flattering toward users. Sycophantic AI has been shown to entrench attitudes, yet users frequently fail to recognize it (a phenomenon we call "sycophancy blindness"). We tested whether increasing users' awareness of sycophancy protects them from its harmful effects in two preregistered experiments (n = 1,590). In the first, participants received a brief written warning about sycophancy before conversing with a sycophantic chatbot. In the second, participants watched a video of a sycophantic AI validating several other users, including users on opposite sides of the same conflict, before interacting with it themselves. Both interventions changed how participants evaluated the AI. The warning reduced the AI's perceived objectivity, and the video reduced enjoyment of the AI --- an effect mediated by the reduced belief that its validation was uniquely earned. We then pooled our experiments with two prior studies of sycophancy awareness interventions (six interventions total, n = 3,982). The pattern across experiments was consistent: while the interventions made the sycophantic AI appear less objective and trustworthy, none reduced its persuasiveness. These results suggest that individual-level interventions, such as warning labels or AI literacy, may not be enough to protect users from AI harms. --- ### OrganLens: Organ-Specific Representation Learning for CT Foundation Models Authors: Zhixuan Ge, Anqi Li, Sadeer Al-Kindi, Hanwen Xu, Wei Qiu Published: 2026-07-28 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2607.25164v1 A CT examination captures multiple organs, but many biomedical questions concern abnormalities, prognosis, or longitudinal change in a specific organ. These questions require a separate representation for each organ within the same CT volume. Existing CT foundation models commonly produce a single volume-level representation, while recent anatomy-aware methods either encode pre-separated organ volumes or explicitly disentangle images into organ token groups. The former may remove clinically relevant surrounding context, while the latter does not condition a shared encoder on a selected organ before its features are formed. We introduce OrganLens for organ-specific representation learning through self-supervision. An organ identity conditions a shared CT encoder, while organ-specific distillation and anatomy-mask supervision shape features for anatomy-weighted pooling into organ-specific representations. At inference, the shared model produces 11 organ-specific representations without external segmentation masks. We evaluate OrganLens on CT-RATE, RAD-ChestCT, INSPECT, and NLST across diverse acquisitions and downstream evaluations. Relative to CT-pretrained DINOv2, heart representations raise CT-RATE cardiomegaly AUROC from 0.910 to 0.953, while lung representations improve the Harrell C-index for NLST lung-cancer mortality by 14.2\%. The global representation reaches INSPECT Recall@10 of 33.09\% and 32.04\% for text-to-image and image-to-text retrieval, respectively. Across organ-related tasks, anatomically matched representations provide stronger task-relevant signal, while the global representation retains broad utility. OrganLens offers a scalable approach to organ-specific CT representation learning with a shared encoder. More broadly, it provides the medical research community with a reusable framework for studying organ-specific disease across cohorts and clinical endpoints. --- ### A Riemannian View on Active Subspaces Authors: Zachary Grey Published: 2026-07-28 Categories: math.NA, cs.LG, math.DG arXiv: https://arxiv.org/abs/2607.25163v1 Active subspaces provide an explainable, eigenvalue-ordered principle for studying how scalar-valued quantities of interest change the most, on average, over a reduced basis of Euclidean domains. Composition with parallel transport generalizes this principle from Euclidean space to quantities of interest defined over Riemannian manifolds, and the resulting intrinsic formulation is contrasted with the extrinsic, embedding-based gradient average of manifold learning. Either strategy is studied in an intrinsically local sense, restricted to mean-centered geodesic-balls, and within that scope the two are not identical: on the central tangent space, eigenvalues agree to second order in the geodesic radius of the sampled domain, while dominant eigenspaces agree at the same order relative to the spectral gap. Extending activity beyond that central space then calls for either recomputed decompositions over changing tangent spaces or, intrinsically, parallel transport of a single central frame. Hyperspheres are emphasized throughout as a particular manifold of interest, motivated by applications over preshape spaces for statistical shape analysis. Numerical examples over the 2-sphere illustrate the formalism, including the derived ridge recovery at a curvature-limited quadratic rate. --- ### PreDiff-LM: Pretrained Discrete Masked Diffusion Language Modeling with Hybrid Attention Authors: Zhengtao Yao, Runhao Li, Xupeng Chen, Jiayi Cheng, Chenqian Le, Michael Yue, Jesson Wang, Siheng Wang, Guang Yang, Haoyan Xu, Chenhao Wei, Zhengqing Yuan et al. Published: 2026-07-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.25157v1 Discrete masked diffusion language models support bidirectional generation and infilling, but adapting pretrained autoregressive (AR) transformers requires reconciling causal pretraining with bidirectional denoising. We study this problem at the level of attention rather than claiming AR-weight reuse itself as novel. PreDiff-LM preserves causal attention within the observed prompt while allowing full bidirectional attention within the masked target. Under a matched GPT-2 Medium, WikiText-103, 90K-step setup, this hybrid mask improves unconditional perplexity from 34.1 to 28.7 and MAUVE from 0.71 to 0.78 over uniform bidirectional attention with the same AR initialization. Attention adaptation also composes with a DiffuGPT-style objective adaptation, reaching 26.9 perplexity. Pretrained initialization reduces the steps required to reach perplexity below 50 from about 350K to 8K, although a compute-matched fine-tuned AR model remains stronger at equal scale (18.9 versus 28.7). Beyond perplexity, PreDiff-LM improves repetition, distributional quality, four zero-shot downstream tasks, and human preference over prior diffusion baselines. The results position hybrid attention as a complementary mechanism for adapting pretrained causal backbones, while making explicit the remaining quality and inference-efficiency gaps to optimized AR models. --- ### Accurate structural modeling of chemically diverse molecular interfaces with Vilya-2 Authors: Vilya Research, :, Pascal Sturmfels, Naozumi Hiranuma, Milad Salem, Benjamin D. Sellers, Stephen Rettie, CJ San Felipe, Chase A. P. Wood, Jeffrey K. Holden, Adam P. Moyer, Patrick J. Salveson et al. Published: 2026-07-28 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.25156v1 Structure-prediction networks built on co-evolutionary statistics have transformed protein-based drug discovery, yet their accuracy does not extend to peptide therapeutics--an increasingly important modality defined by non-canonical residues, macrocyclization, and complex topologies. We introduce Vilya-2, a diffusion transformer that extends the all-atom representation of Vilya-1 from modeling individual molecules to modeling their interactions with protein targets. This all-atom representation enables transfer learning between different molecular types, and delivers highly accurate structural modeling of peptides across sizes, classes, and compositions bound to therapeutically relevant targets. By generating diverse structural ensembles and ranking them with calibrated confidence, Vilya-2 recovers 59.1% of peptide interfaces to sub-2 Å backbone RMSD, far exceeding the performance of a representative co-folding model even when that model is given the bound receptor as a template. In addition, Vilya-2 is state-of-the-art at small-molecule docking, and generalizes to novel protein-small molecule complexes unlike those seen in training. It also generalizes to modeling molecular conformations of diverse macrocycles and disulfide-stapled miniproteins several-fold larger than any molecule seen in training. Finally, Vilya-2 can be used as a foundation model, and fine-tuned to enrich for active compounds in hit-to-lead campaigns. By unifying predictive accuracy with broad generalizability across chemical space, Vilya-2 is the structure-prediction oracle that de novo peptide design pipelines require--establishing the all-atom approach as a general foundation for the design and evaluation of de novo peptide therapeutics. --- ### Do Synthetic Personas Predict Real Audience Response? A Sim-to-Real Study Where a No-Persona Baseline Beats Persona-Based Copy Simulation Authors: Alexandre Cristovão Maiorano Published: 2026-07-27 Categories: cs.AI, cs.CL, cs.CY arXiv: https://arxiv.org/abs/2609.25010v1 Marketers increasingly use large language models (LLMs) as "synthetic personas" to predict how an audience will react to a piece of copy before it ships, encouraged by evidence that profile-conditioned LLMs mimic human samples. But is that prediction actually valid against real behaviour - and does the persona machinery help? We present a sim-to-real validity study using the Upworthy Research Archive - thousands of headline A/B tests on shared real traffic, with measured click-through - as held-out ground truth. We compare a ten-persona panel, grounded in the real audience's demographics, against a no-persona zero-shot baseline that simply asks the model how likely a typical reader is to click. Two findings stand out. First, ground-truth reliability is the binding constraint: most A/B tests have no statistically distinguishable winner, so validity can only be measured on the reliable subset (n = 399). Second, and counter to the persona-simulation premise, persona conditioning degrades predictive validity: the no-persona baseline ranks variants markedly better (Kendall τ = 0.361, a medium effect; top-1 accuracy 49.2%) than the persona panel (τ = 0.084; top-1 34.6%), with non-overlapping confidence intervals. Asking the model directly taps an accurate population-level prior; forcing it to role-play specific personas injects bias and noise. The result replicates across three independent Upworthy splits, holds in direction on a different-domain news dataset, and is robust to seed, prompt phrasing, and model choice - across three Gemini tiers and a different model family (OpenAI gpt-4.1, significant paired gap). The takeaway: for predicting aggregate engagement, a plain LLM ranker beats persona simulation - synthetic personas are not merely a weak predictor, they are worse than not using them. All numbers regenerate from a public, artifact-first replication package. --- ### Same Quantity, Different Answer: Numerical Representation Invariance in Language Models Authors: Ephraim Atta-Duncan Published: 2026-07-27 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2609.25009v1 Numerically equivalent word problems should yield the same canonical answer whether a quantity is written as a decimal, fraction, percentage, number word, scientific notation, or an exactly converted unit. We generate 3,600 exact-rational problems and 8,600 prompts spanning five identity-preserving transformation families, and evaluate five open-weight systems. After a fixed syntax audit that normalizes common answer forms without an LLM judge, canonical accuracy is 0.969-0.996, but orbit correctness falls to 0.848-0.981 and orbit invariance to 0.851-0.981; invariant-but-wrong orbits account for at most 0.003. Most of the broad strict-parser collapse arises because multiplication-form scientific notation lies outside the implemented number grammar, illustrating how evaluator interfaces can masquerade as reasoning failures. A distinct semantic pathology remains: Mistral Small 4 scores 0.699 on unit-converted inputs and produces 265 errors differing from the label by exact powers of ten. In a separate 9,000-call experiment that allocates equal calls to the compared arms, representation consensus does not outperform paraphrase consensus on a low-error subset and produces substantially more false alarms. The accompanying ancillary archive contains the frozen benchmark, evaluation and audit records, consensus raw responses, manifests, analysis code, and a one-command paper build. --- ### Training a Language Model End-to-End in Rust: An Experience Report Authors: Arif Adito Published: 2026-07-27 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.25008v1 I pretrained a language model end-to-end in Rust - alone, with no team, no PyTorch, and no Python in the training path - for $164 in rented GPU time. I report that as an achievement, not a recommendation: the more useful contribution is a measured failure taxonomy of the two leading Rust ML frameworks, Candle and Burn, as training (not inference) backends in 2026. I document five Candle defects, including fused kernels that silently produce no gradient, and three Burn defects, including a backward pass at roughly 3% of theoretical GPU throughput and a kernel-fusion path that segfaults mid-training at multi-billion-parameter scale. Every one passed ordinary loss-curve inspection; none announced itself. I describe the verification discipline that caught six such silent failures, centered on a gradient-flow arbiter: a test that runs one forward/backward pass and asserts every trainable parameter receives a finite, nonzero gradient, generalizable to any framework. The trained model (roughly 0.4B parameters, Bangla-first) shows strong Bangla language-modeling signal - a per-token negative log-likelihood of 0.93 against 12.60 for a random-initialized twin - while scoring at chance on English commonsense multiple-choice, the expected outcome of a deliberately small, Bangla-weighted budget (about 2 billion tokens, 54.6 hours, one rented H100). I also report a tokenizer-fertility trap in Bengali script: naive byte-level tokenization collapsed Bangla to roughly 1.4 characters per token against English's 3.9, silently inverting the corpus's language balance; fixing it reached roughly 4.1. To my knowledge, this is among the first documented end-to-end LM pretraining runs in pure Rust. After this run I moved training to PyTorch and kept Rust for on-device serving: in my hands, Rust is not yet a competitive place to train a language model, though it may be a good place to serve one. --- ### The Public Discourse Corpus (PDC): A Speaker-Attributed Dataset for Valence and Epistemic Modality with Target Speaker Participation Authors: Bo Chen Published: 2026-07-27 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.20232v1 We introduce the \textbf{Public Discourse Corpus (PDC)}, the first dataset of public-figure interview speech jointly annotated for affective valence and epistemic modality. The corpus contains 998 videos from 100 speakers across seven professional domains, yielding 186,642 sentences (3.1 million words) after sentence segmentation and filtering. To ensure that all retained videos contain analyzable speech from the intended speaker, we introduce \textbf{Target Speaker Participation (TSP)}---a five-category annotation taxonomy with documented inter-annotator reliability ($κ= 0.616$)---as a key methodological contribution that any corpus construction project can adopt. Target-speaker turns are separated from interviewer and third-party speech through an \textbf{audio-first diarization pipeline} combining local Whisper ASR with pyannote speaker separation, released as an open-source implementation. We release the annotated corpus, the annotation tools, the cross-provider validation sample, and the complete processing pipeline. The dataset is available at https://huggingface.co/datasets/ictchenbo/public-discourse-corpus. --- ### Self-complementary completions on six vertices Authors: Xinan Dai, Wenhao Deng, Yingdong Shi, Tailin Wu, Yuchen Yang Published: 2026-07-27 Categories: math.CO, cs.AI arXiv: https://arxiv.org/abs/2609.20231v1 Let \(\cthreshold(n)\) be the largest integer \(q\) such that every loopless digraph on \(n\) vertices with at most \(q\) arcs is isomorphic to a spanning subdigraph of a self-complementary digraph of order \(n\). We prove that \(\cthreshold(6)=7\). The upper bound is witnessed by \( \bK{3}\dunion (x\longrightarrow y\longrightarrow z), \) and follows from a direct argument with a self-complementing permutation. We also determine the complete eight-arc obstruction layer: it consists of five isomorphism classes, or three after converse digraphs are identified. All five are arc-minimal. Each nevertheless packs with an isomorphic copy of itself, so ordinary packing is strictly weaker than same-order self-complementary completion already at this first failure layer. --- ### Independence-System Realisations in Single-Source Unsplittable Flow Authors: Koyar Afrasyab Published: 2026-07-27 Categories: cs.DS, cs.AI arXiv: https://arxiv.org/abs/2609.17568v1 Additive-congestion constraints in single-source unsplittable flow can enforce stable-set structure. This note isolates and generalises that mechanism. We introduce a path-closed notion of realising an independence system by the zero-cost choices of primary terminals in a directed acyclic flow instance. The definition quantifies over every directed source-terminal path and therefore remains valid under prefix borrowing, suffix splicing, and hybrid routes.Our main result extends the triangle mechanism: every finite loopless independence system has a polynomial-size realisation, measured in the incidence size of its minimal forbidden sets. Hence every finite simple graph, and more generally every hypergraph independence system without singleton forbidden hyperedges, is representable by an acyclic single-source gadget. We then specialise the construction to odd cycles. For C_{2k+1}, a uniform rational family produces a fractional cheap-selection vector that violates the odd-cycle inequality. A potential shift converts a signed connector separator into nonnegative arc costs and gives the exact cost-preserving additive-congestion threshold tau = 1 - bq. Within the symmetric family, the supremum threshold is (k+2)/(2(k+1)), which tends to 1/2. For C5, an exact certificate independently derives all source-terminal paths and enumerates all 3^10 = 59049 unsplittable routings using rational arithmetic. --- ### Algorithm Validation as a Policy Audit: Evidence from Race-blind Charging Authors: Muskan Walia, Joe Nudell, Alex Chohlas-Wood Published: 2026-07-27 Categories: cs.CY, cs.CL arXiv: https://arxiv.org/abs/2609.13174v1 California recently required all prosecutors in the state to conduct a "race-blind charging" decision by reviewing case documents in which selected race-related proxies have been redacted. We validate bc2, an open-source, LLM-based algorithm that we developed to automate this redaction and that was used to facilitate race-blind review in more than 119,000 real-world cases in 2025. We evaluate two distinct questions: whether bc2 faithfully implements the state's requirements and whether those requirements, even when faithfully implemented, advance the goal of race-blind decision-making. To do so, we draw on a corpus of nearly 5,000 real-world police reports that we assembled from jurisdictions across the United States. Under a stringent document-level measure, we find that the latest version of bc2 faithfully implements the legal mandate on 96.7% of narratives in our sample. This performance represents a substantial improvement over earlier versions of bc2 and exceeds that of leading open-source redaction methods. Our validation also shows that California's mandate misses key proxies for race, including location information. Redacting these additional proxies beyond those covered by the state mandate, as bc2 does, eliminates 43.1% of the predictive signal that remains after compliance with the mandate. These findings show that validation can do more than assess technical compliance: it can also improve algorithms and help policymakers achieve underlying policy goals. --- ### Pixel Decodability Is Not a Compression Signal: Causally Evaluating Importance Proxies for Visual KV-Cache Eviction Authors: Chenyu Zhou, Qiliang Jiang, Shuning Wu, Xu Zhou Published: 2026-07-27 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.13012v1 Vision-language models retain a substantial amount of pixel-decodable visual content in their visual key-value cache. We show, in our setting, that this retention is task-inert: across our preregistered tests, how much a unit retains never positively tracks whether the computation that answers the question causally relies on it. We measure retention with a learned pixel-inversion decoder and causal use with single-super-patch KV ablation, the teacher-forced drop in gold-answer log-probability, and relate the two within images under a preregistered, sign-calibrated, held-out design. Retention is decoupled from attention and, in a well-powered null, from causal utilization. Utilization is not inert to every proxy: attention weakly but significantly tracks it, the only signal we find that does and the design's positive control. We characterize pixel-decodable retention as an informational axis of the visual KV cache, orthogonal to the functional one. How much task-inert content a cache holds differs by architecture in our model pair: the encoder-free model retains 2.7 times more than the encoder-based one. The engineering consequence is a controlled negative result. At super-patch granularity, deconfounded pixel-decodable retention ranks KV eviction no better than random; at token granularity it acquires only a weak inverse-importance signal at larger budgets, dominated at every budget by attention magnitude. In our setting, pixel-decodable reconstructability is not a competitive KV-compression signal at any granularity we test. --- ### Lost in Compression: A Controlled Cross-Lingual Audit of Extractive Prompt Compressors Authors: Mantas Lukauskas Published: 2026-07-27 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2608.26175v1 Extractive prompt compression promises to cut LLM inference costs by removing low-information tokens, and learned compressors such as LLMLingua-2 report strong results on English benchmarks. Most other languages already pay a token premium: the same content costs 1.3-1.8x more tokens than in English. We ask whether compression closes or widens this gap. Using fully parallel data in ten languages spanning five scripts, with controls budget-matched in the target model's tokenizer, we audit four learned compressors against four deterministic baselines, on eleven target models from ten vendors (over 250,000 evaluation calls). Three of the compressors are trained with English supervision (LLMLingua-2 XLM-R/mBERT; Kompress-v2 from the production Headroom stack); the fourth, XProvence, is trained multilingually. First, the transfer gap is real, replicates across target models and compressor backbones, and is strongly rate-dependent: at a 0.33 keep-rate English retains 57-62% of normalized context utilization while Lithuanian retains 10-24% and Chinese essentially none, despite Chinese having the smallest token premium. Second, the gap tracks compression supervision data, not architecture. All three English-trained compressors show it, deterministic methods show no comparable gap, and the multilingually trained XProvence v1 shows none. Its v2 release, retrained on translated data, empties 92% of Chinese contexts at its aggressive threshold without any warning. Third, in a harder long-context setting, aggressive learned compression drives compressed contexts to or below no-context utility in three of five non-English languages. A translate-then-compress pipeline matches or beats native compression at roughly half the token cost in three of five tested languages. We release all code, compressions, and model outputs. Safe compression budgets are much smaller outside English. --- ### Metaplasticity as adaptive gradient preconditioning for incremental learning Authors: Isabelle Aguilar, Zayn Andre Zainal, Omid Kavehei Published: 2026-07-27 Categories: cs.LG arXiv: https://arxiv.org/abs/2608.14634v1 Biological intelligence naturally prevents catastrophic forgetting through Complementary Learning Systems (CLS) theory, a macroscopic consolidation process driven at the local level by synaptic metaplasticity: the continuous, history-dependent neuromodulation of individual synapses. While artificial neural networks struggle with the stability-plasticity dilemma in non-stationary environments, existing solutions often require task labels or incur massive memory overhead, diverging from biological reality. Re-framing this localized neuromodulation as an optimization-driven process, we introduce $\textbf{SynGAP}$: $\textbf{Syn}$aptic $\textbf{G}$eometric $\textbf{A}$daptive $\textbf{P}$reconditioning. SynGAP is a task-free continual learning framework based on adaptive gradient preconditioning. Rather than relying on explicit episodic triggers, SynGAP simulates real-time metaplasticity by maintaining an exponential moving average of the Fisher Information Matrix over a continuous data stream. During the optimization step, these dynamic metaplastic states are translated into a bounded multiplicative mask that preconditions raw gradients, selectively attenuating updates to critical historical parameters. Empirical evaluations demonstrate SynGAP's superior ability to mitigate catastrophic forgetting compared to established baselines. On the Split CIFAR-100 benchmark, SynGAP delivers a $4\times$ increase in accuracy compared to EWC++ and outperforms Experience Replay (ER) by almost $10\%$, while reducing the forgetting measure by over $10\%$ against both methods. Furthermore, on the CORe50 benchmark, SynGAP achieves about $68\%$, a $10\%$ improvement over optimizer baselines. By mathematically formalizing continuous biological metaplasticity as stable gradient-based regularization, SynGAP offers a highly robust and memory-efficient solution for adaptive intelligence at the edge. --- ### Wolff-Parkinson-White Detection at 471:1 Class Imbalance: A Leakage-Controlled Study of the Data Bottleneck Authors: Nathael Altman Published: 2026-07-27 Categories: eess.SP, cs.LG arXiv: https://arxiv.org/abs/2608.14633v1 Wolff-Parkinson-White (WPW) syndrome is a congenital cardiac pre-excitation, clinically important and often missed on the resting 12-lead ECG. Detection is hard: the signature is subtle and the condition rare. We pool two public 12-lead corpora, PTB-XL and Chapman-Shaoxing-Ningbo: 66,951 recordings, 142 of them WPW, a prevalence of 0.21% (about 471:1). Under one pre-specified, leakage-controlled protocol, with a held-out fold contacted exactly once, we compare seven representations of the signal, holding the split and the evaluation fixed. Within these corpora and under a modest compute budget, added diversity and capacity do not raise the ceiling: the most orthogonal detector significantly hurts, a feature-union model matches a two-member vote, a convolutional network reaches the wavelet detector without exceeding it, and self-supervised pretraining fails a pre-specified gate. A leak-free learning curve, re-selecting features at every size, still rises at the full 115 positives for the strongest deployed detector (paired 90-to-100% difference +0.027, 95% CI (0.019, 0.033)), so it is not shown to have saturated. An error analysis tested against independent evidence finds that the missed cases have a narrower QRS, confirmed by an on-machine measurement outside our pipeline after we show the sign of this effect depends on which delineator measures it; that uncertain labels show no enrichment among the misses; and that some apparent false positives are recordings the corpus itself codes as pre-excited, placing part of the label problem in the negative class. We measure the optimism of non-nested selection at 0.11 to 0.13 average precision. The deployed output is a percentile rank in a frozen reference distribution, not a probability. On the held-out fold, on 14 positives, it reaches an average precision of 0.595 and an ROC area of 0.950. It is a screening pre-filter, not a diagnostic tool. ---