# Research Papers — 2026-06 · bundle page 95 of 260 > Full-text markdown of papers #4701–4750 of 12993 published in 2026-06, newest first. AI-readable mirrors of the full 12993-paper 2026-06 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-06/page/95 [← Previous bundle](https://aipod.dev/papers/archive/2026-06/page/94.md) [Next bundle →](https://aipod.dev/papers/archive/2026-06/page/96.md) ### Neurosymbolic Clinical Trial Matching via LLM-Driven Abduction and Logical Verification Authors: Baiyang Qu, Leonardo Ranaldi, Xi Wang, Marco Valentino Published: 2026-06-18 Categories: cs.AI arXiv: https://arxiv.org/abs/2606.20895v1 Large Language Models (LLMs) offer a promising path to automate Clinical Trial Matching (CTM), but still struggle with the deterministic verification required for complex eligibility criteria. Conversely, purely symbolic methods provide formal rigour but break down when faced with incomplete patient records and noisy clinical evidence. To bridge this gap, we investigate a hybrid framework for CTM combining LLMs with logical verification. In particular, we introduce an abductive neurosymbolic CTM framework (αNeSy-CTM), which leverages the linguistic and world knowledge in LLMs to support reasoning over noisy and underspecified clinical text. Extensive evaluation demonstrates that αNeSy-CTM substantially outperforms standalone LLM baselines, achieving up to 30% relative improvement over zero-shot baselines. In addition, our analyses confirm the impact of abductive reasoning on CTM, with αNeSy-CTM exhibiting improved accuracy, specificity, and robustness over a non-abductive neurosymbolic setting. Furthermore, αNeSy-CTM and Chain-of-Thought (CoT) reasoning prove highly complementary, highlighting the potential for a hybrid routing policy. Ultimately, this paper demonstrates the impact of neurosymbolic methods for automating CTM, providing a path toward the next generation of auditable, LLM-driven clinical applications. --- ### Exploiting Neural Audio Codec Latents for Adversarial Audio Attacks Authors: Sameek Bhattacharya, Bharath Krishnamurthy, Ajita Rattani Published: 2026-06-18 Categories: cs.SD, cs.AI, cs.CR arXiv: https://arxiv.org/abs/2606.20893v1 Deep learning-based audio classification systems, including automatic speaker verification, are vulnerable to adversarial attacks. Realistic real-time threat assessment remains difficult because optimization-based methods, such as projected gradient descent (PGD) and Carlini-Wagner, require costly iterative updates in the high-dimensional waveform domain. Generative attacks allow single-shot synthesis but often introduce perceptible artifacts or depend on computationally intensive architectures, while diffusion and autoregressive approaches incur high inference latency. To address this gap, we propose a generative attack framework operating in the continuous latent space of a neural audio codec. A conditional generator synthesizes class-specific perturbations in a single forward pass and decodes them into adversarial waveforms. Our method achieves targeted attack success rates up to 99% with sub-7 ms inference, outperforming generative baselines while reducing latency by 24x. --- ### Go-with-the-Track: Video Compositing and Motion Control with Point Tracking Authors: Koichi Namekata, Yash Kant, Zhizheng Liu, Ryan D Burgert, Yuancheng Xu, Kuan Heng Lin, Emmett Steven, Julien Philip, Li Ma, Andrea Vedaldi, Paul Debevec, Ning Yu Published: 2026-06-18 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2606.20891v1 Filmmaking demands precise motion control and reference image compositing -- capabilities that existing methods treat separately. Point-track-conditioned image-to-video models restrict content insertion to the first frame, while reference-to-video models lack fine-grained spatial-temporal control over how reference content integrates across frames. We present Go-with-the-Track, which unifies both capabilities by jointly conditioning on multiple reference images and reference-anchored point-tracks -- extending conventional point-tracks to explicitly establish correspondences between generated frames and reference images, thus enabling precise compositing and motion control throughout the video. To achieve this, we introduce spatially-aware point-track embeddings that encode the full sequence of point-track coordinates using a coordinate-wise MLP followed by temporal pooling. This representation captures the spatial characteristics of each point-track (serving as a unique identifier), while the embedding similarity correlates directly with spatial proximity, enhancing the model's ability to distinguish and associate point-tracks. We inject these point-track embeddings into a video diffusion transformer via a lightweight adapter, resolving the pixel-to-patch resolution mismatch while avoiding the substantial motion detail loss inherent in naive point-track subsampling. We use a hybrid training strategy to train jointly on dynamic, static, and synthetic scene video datasets to boost motion controllability. Experiments demonstrate that Go-with-the-Track achieves superior motion and reference control in a single model and enables new capabilities: multi-reference conditioned video generation with point-track driven compositing, as well as camera control for both static and dynamic scenes. Project Page: https://eyeline-labs.github.io/Go-with-the-Track/ --- ### Topic-to-Timestamp Alignment by Constrained Evidence Selection Authors: Zeynep Yılbırt, Marina Litvak, Michael Färber Published: 2026-06-18 Categories: cs.CL, cs.IR arXiv: https://arxiv.org/abs/2606.20890v1 Meeting archives are difficult to search when users remember what was discussed but not when. We study topic-to-timestamp alignment: given a natural-language topic and a timestamped meeting transcript, the goal is to return the time at which the topic is discussed. A standard RAG setup can retrieve relevant transcript excerpts, but still asks the language model to generate a timestamp, which can produce unsupported or invalid timecodes. We therefore recast timestamp prediction as constrained temporal candidate selection: the system retrieves timestamped transcript chunks, and the model selects the candidate that best grounds the topic instead of generating a timecode. On 420 topic-timestamp queries from 200 municipal meeting transcripts, this increases Recall@5 from 31.9% to 50.0%, reduces MAE from 837.0 seconds to 761.0 seconds with Mistral-7B-Instruct, and increases the number of parseable outputs from 373 to 419 of 420 queries. The results suggest that temporal grounding in long transcripts depends strongly on retrieval quality and output design, not only on the choice of the language model. --- ### Temporal Causal Prior-Data Fitted Networks for Panel Data with Learned Reliability Signals Authors: Shravan Talupula, Saurabh Sharma Published: 2026-06-18 Categories: cs.LG, stat.ML arXiv: https://arxiv.org/abs/2606.20889v1 Estimating causal effects in industrial time series requires handling temporal dynamics, time-varying treatments, and unobserved confounders. Existing causal foundation models (CausalPFN, CausalFM) operate only on static cross-sectional data; neural temporal methods (CRN, G-Net) require per-dataset training; and concurrent temporal-PFN proposals have not been demonstrated at industrial scale. None output explicit per-pair reliability signals alongside their CATE estimates. We introduce Temporal Causal Prior-Data Fitted Networks (TCPFN), a foundation model for zero-shot temporal causal discovery with learned reliability signals. TCPFN makes four contributions: (1) a Causal Judgment Head that jointly predicts null-effect probability, confounding strength, identifiability, mediation fraction, and causal regime; (2) a mixed training prior covering six causal regimes (independent, direct, confounded, mediated, time-varying confounded, feedback) plus CausalFM-style front-door and instrumental-variable priors; (3) a discrete-token panel-data architecture with cross-attention masking that prevents inter-horizon leakage; (4) zero-shot inference at industrial scale via FAISS-based context selection and one-step posterior correction. On 19 benchmark datasets across five domains, TCPFN achieves competitive zero-shot causal discovery: AUROC 0.96 on Tennessee Eastman, 0.93 on SWaT, 0.98 on Causal Rivers, 0.97 on CAUSRCA. The null detector reaches NullF1 0.94, AUROC 0.99. TCPFN scales to V=1,275 on a proprietary Kraft pulp-and-paper dataset in 6 hours on a single GPU; PCMCI, a CPU-only library, on a V=666 sub-panel of the same data took 81.5 hours, extrapolating by O(V^2) to ~12.5 days at V=1,275. TCPFN's top edges identify cross-subsystem causal relationships while PCMCI's surface within-instrument controller-measurement coupling -- a scalability case study. --- ### Fine-grained Human Motion Understanding with Language Models Authors: Thomas Markhorst, Zhi-Yi Lin, Jouh Yeong Chew, Jan van Gemert, Xucong Zhang Published: 2026-06-18 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.20888v1 In this work, we propose \methodname, an LLM-based model for fine-grained human motion understanding that represents motion as a sequence of skeletal poses with explicit timestamps for each pose. Each pose encodes body joint positions and is temporally grounded with timestamp tokens, allowing the model to reason about motion order, duration, and rhythm. To study what supervision is needed for motion-language reasoning, we construct a diverse training mixture spanning pose captioning, pose question answering, motion captioning, and motion question answering. Our ablations show that the primary gains come from the diversity of pose- and motion-level supervision, while staged training provides a smaller additional benefit. Different from previous works that rely on ground-truth 3D motion capture, our approach supports both 2D and 3D skeletal motion representations through a unified pose encoder, and can optionally incorporate video to provide contextual information. Extensive experiments on BABEL-QA, HuMMan-QA, CompMo, NTU-RGB+D, and QEVD-Coach demonstrate that our method achieves state-of-the-art performance across multiple benchmarks, highlighting the effectiveness of explicit temporal encoding and diverse pose- and motion-level supervision for fine-grained human motion understanding. Notably, even when using only 2D skeletal input, our approach surpasses previous 3D-based methods. --- ### Toward Parking Spot Occupancy Recognition: A Self-Supervised Approach Authors: Luan Marko Kujavski, Rayson Laroca, Paulo Lisboa de Almeida Published: 2026-06-18 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.20886v1 As urban areas expand, automatic monitoring of parking lots becomes essential for efficient and sustainable cities. This work proposes a self-supervised approach for parking spot occupancy recognition that requires no labeled samples from the target parking lot. Building upon a self-supervised transfer learning fine-tuning protocol, the proposed training strategy consists of two self-supervised stages: first on unlabeled generic data and then on unlabeled target-specific data, followed by supervised fine-tuning using only generic parking lot labels. We adopt SimCLR with a ResNet-50 encoder and evaluate the method under a leave-one-out cross-environment protocol on three public datasets: PKLot, CNRPark-EXT, and PLds. We also introduce a two-stage deployment strategy in which a Strong General Model is initially deployed, followed by a Specialized Model that incorporates unlabeled images collected during the first N days of deployment in a self-supervised manner. Experimental results show that the Strong General Model alone outperforms supervised and self-supervised baselines, achieving an average accuracy of 97.2%, which further improves to 97.8% with the proposed two-stage strategy. These results demonstrate that self-supervised learning enables a scalable and labelefficient solution for real-world parking occupancy monitoring. Our trained models and source code are publicly available at https://github.com/LoanMaikon/Parking-Spot-Occupancy-Recognition. --- ### The Substrate Collapse: AI Code Generation Invalidates Authorship-Based Knowledge Metrics Authors: Brett Wheeler Published: 2026-06-18 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2606.20882v1 Software engineering has long inferred where a system's knowledge resides from who authored its code. The truck factor, the Degree-of-Authorship metric, and the degree-of-knowledge model all rest on one inference -- that authoring a region of code is evidence of understanding it -- and for most of software's history it was a workable proxy, because code entered a repository only when a human wrote it, which forced at least transient understanding. This paper argues that AI code generation severs that inference at its root, and that the consequence is not the degradation of the authorship-based metrics but their invalidation as a class. When an agent generates a module and a human merges it, the version-control record still attributes authorship, but the attribution no longer licenses any conclusion about comprehension: the same footprint is now compatible with full, partial, or no understanding. The metric still returns a number; the number measures a substrate that has come uncoupled from the quantity it was used to estimate. The collapse is corroborated by the field's own measurement failures, and the methodological corollary is load-bearing: the instrument the comprehension-debt era needs cannot be built by refining the knowledge-concentration metrics, because no function of an authorship footprint recovers an inference the footprint no longer supports. The replacement must be grounded in evidence of comprehension rather than authorship. I state a falsifiable prediction that discriminates the two -- that systems with a healthy authorship-derived truck factor but low comprehension-measured retention will suffer incident-resolution failures the authorship metric does not predict -- and argue that building the comprehension-grounded instrument at the scale of a system and a team is the field's open measurement problem, left open here. --- ### When Do Intrinsic Rewards Work for Code Reasoning? A Comprehensive Study Authors: Xiaolong Jin, Xuandong Zhao, Wenbo Guo, Xiangyu Zhang, Dawn Song Published: 2026-06-18 Categories: cs.AI arXiv: https://arxiv.org/abs/2606.20881v1 Reinforcement learning with verifiable rewards (RLVR) has driven substantial progress in large language model reasoning, but relies on ground-truth supervision that is costly or infeasible, especially in coding tasks. Recent work addresses this by deriving rewards from a model's own signals, such as majority voting or confidence-based scores, achieving notable success on mathematical reasoning benchmarks. However, code generation poses distinct challenges: programs are structurally complex, semantically equivalent solutions may differ syntactically, and verification typically requires execution. Whether these intrinsic reward methods transfer effectively to code remains unexplored. In this work, we present a systematic empirical study of intrinsic reward methods for code generation. We conduct extensive experiments on LiveCodeBench, systematically evaluating representative certainty-based Reinforcement Learning from Internal Feedback (RLIF) approaches under different training scenarios and hyperparameter settings. Our experiments reveal that certainty-based methods yield early gains but inevitably collapse: models progressively shorten outputs and lose reasoning capability, with collapse speed sensitive to sample size and temperature. When used to initialize RLVR training, RLIF pre-training offers no significant improvement over training from scratch. We also provide actionable recommendations for using intrinsic rewards for training code reasoning models. Our study shows both the promise and limitations of intrinsic reward methods for code, informing future work on code models and agents. --- ### Adversarial observations in probabilistic State-Space Models for robust Reinforcement Learning Authors: M. Santos-Pascual, D. Ríos Insua Published: 2026-06-18 Categories: stat.ML, cs.LG, stat.ME arXiv: https://arxiv.org/abs/2606.20880v2 Decision-making under partial or adversarial observability requires accurate inference of the environment's latent state and its associated uncertainty. This work analyses adversarial attacks on linear state-space models, where the attacker alters observations subject to likelihood constraints that ensure that the perturbations remain statistically consistent with the observation model. We analyse how such adversarial yet plausible observations shift inference about latent states and affect downstream decision-making and the performance of reinforcement learning agents. In addition, we introduce an online Bayesian defence based on directional covariance adaptation, which selectively reduces the influence of observations by comparing their estimated impact with that of the computed most disruptive direction, while preserving information in the remaining orthogonal observation subspace. The proposed framework provides a principled approach to constructing robust inference and decision-making systems, with direct relevance to safety-critical applications such as robotics, where reliable operation under sensor noise, partial failures, and adversarial conditions is essential. --- ### Understanding Latent Flow Models for Tabular Data Synthesis: Targets, Paths, and Sampling Authors: Bahrul Ilmi Nasution Published: 2026-06-18 Categories: cs.LG, stat.ML arXiv: https://arxiv.org/abs/2606.20878v1 Synthetic tabular data enables microdata sharing in regulated domains, yet deploying continuous-time generative models requires balancing analytical utility, disclosure risk, and computational cost. Latent-space flow models are flexible, but theoretical equivalences across learning targets, probability paths, and sampling dynamics can translate into different behaviour under finite-step integration and explicit compute budgets. We present an empirical study of tabular latent flow models across seven datasets, evaluating velocity, score, noise, and posterior matching objectives under optimal transport (OT) and variance-preserving (VP) paths, ODE and SDE sampling, and varying integration budgets. Our contributions are threefold: (1) we show that the learning target largely determines the utility-risk operating regime, with velocity and posterior matching tending to yield higher utility, while score and noise matching tend to achieve lower disclosure risk; (2) we demonstrate that configuration and sampling choices shift performance, with midpoint often improving distributional fidelity and OT paths often tolerating earlier stopping than VP, enabling compute savings under fixed budgets or risk thresholds; and (3) we distil these findings into actionable defaults and practical configuration guidance to support pre-release model selection under disclosure risk and resource constraints. The code implementation and supplementary materials can be accessed in https://github.com/rulnasution/tabular-latent-flow/. --- ### Artificial collectives of specialists and generalists excel at different tasks Authors: John Meluso, Laurent Hébert-Dufresne, Christoph Riedl, H. Oliver Gao Published: 2026-06-18 Categories: cs.MA, cs.AI, cs.SI, nlin.AO, physics.soc-ph arXiv: https://arxiv.org/abs/2606.20877v1 Collective artificial intelligence, where multiple agents work on shared tasks, holds potential to solve expansive problems in fields from medicine to collective governance. But while prescriptive engineering solutions abound, we lack descriptive scientific understanding of artificial collectives, and therefore principles for how to design resource efficient multi-agent systems. Through systematic experiments with optimizing agents, we characterize how agent interpretive abilities, rationality bounds, and task qualities interact to shape collective performance. Agents range from specialists, with narrow interpretive abilities, to generalists, with broad ones. Collectives of specialists correspond to sparse, centralized networks, while collectives of generalists correspond to dense, decentralized ones. We show that interpretive network properties have small performance effects on average (0.07 standard deviations of performance). However, for specific task qualities, these effects are 4.5 times larger (0.33 sd) and can reach much higher for certain task qualities (1.84 sd). This leads collectives of generalists to perform better on tasks that involve generating, choosing, and coordinating, while collectives of specialists with a few generalist mediators perform better on tasks that involve negotiating. Rationality bounds then moderate these relationships. At loose bounds, specialists outperform generalists through more effective sampling of high-dimensional decision spaces. At tight bounds, generalists outperform specialists through better gradient estimation. A fundamental trade-off between performance and convergence speed emerges at moderate bounds. These findings suggest that multi-agent design could benefit from matching interpretive networks to both task demands and agents' computational limits, with implications for the efficiency and energy costs of multi-agent systems. --- ### Machine Learning Classification of Cryopathy Syndromes: A Comprehensive Comparative Study Authors: Nataliya Shakhovska, Valentyna Chopyak, Ivan Izonin, Vira Haievska Published: 2026-06-18 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2606.20874v1 Cryopathy syndromes are difficult to classify because laboratory patterns often overlap across diagnostic categories, while some diagnoses are rare. This makes routine interpretation of cryoglobulin-related tests challenging and increases dependence on expert judgment. The aim of this study was to develop and compare machine learning approaches for automated classification of cryopathy syndromes from laboratory data and to identify a practical strategy for clinical decision support. Methods: We analysed laboratory records from 2,686 patients assigned to 14 diagnostic categories. The dataset included demographic variables, cryoglobulin measurements, precipitation tests, and hemagglutinin and hemolysin titers. Data preprocessing included cleaning, encoding, imputation, normalization, and construction of clinically informed interaction features. We evaluated 12 modelling strategies, including Random Forest, Gradient Boosted Trees, Multi-Layer Perceptron, soft-voting ensembles, class balancing with Synthetic Minority Over-sampling Technique, hierarchical classification, period-aware models, targeted binary classifiers, and probability calibration. Performance was assessed using stratified train-test evaluation and stratified 5-fold cross-validation. The main metrics were macro-averaged F1 score, accuracy, Top-3 accuracy, and expected calibration error. The overall task proved difficult because of marked class imbalance and clinical overlap between diagnoses. The best multiclass performance was achieved by a soft-voting ensemble of Random Forest and Gradient Boosted Trees. Cross-validation confirmed stable performance for the balanced Random Forest model. Tree-based methods consistently outperformed the neural network model. Feature engineering improved discrimination, and the most informative predictors were derived cryoglobulin-based interaction features. --- ### SciLens: Multi-modal Scientific Claim Verification with Agentic Entailment and Grounding Authors: Yueming Wang, Tianshi Zheng, Jiaxin Bai, Yangqiu Song, Ginny Wong, Simon See Published: 2026-06-18 Categories: cs.CL arXiv: https://arxiv.org/abs/2606.20873v1 Scientific discovery increasingly relies on automated systems that generate hypotheses, inspect multimodal evidence, and validate claims at scale. Yet scientific claim verification is not well served by asking a vision-language model for a direct binary judgment: claims often combine numerical results, comparisons, scope qualifiers, and explanatory context, while evidence is encoded in tables and figures with distinct grounding structures. We present SciLens, an evidence-conditioned atomic entailment framework for multimodal scientific claim verification. SciLens decomposes each claim into central empirical atoms and background atoms, grounds the central atoms to modality-specific evidence witnesses, and predicts the final label with an atom-level entailment rule. For tables, atoms are grounded to rows, columns, cells, arithmetic relations, and table scope; for figures, they are grounded through panels, axes, legends, visual encodings, categories, trends, ranks, and qualifier checks. This yields a unified validation procedure in which a claim is supported only if every central empirical atom is entailed by the current evidence. On the SciClaimEval development set, SciLens achieves 79.2% macro-F1 and 63.1% pair accuracy, showing that structured agentic validation improves both evidence sensitivity and interpretability. --- ### A3C3: AI Algorithm and Accelerator Co-design, Co-search, and Co-generation Authors: Selin Yildirim, Yingbing Huang, Deming Chen Published: 2026-06-18 Categories: cs.AR, cs.AI arXiv: https://arxiv.org/abs/2606.20869v2 We present a holistic methodology for artificial intelligence algorithm and accelerator co-design, co-search, and co-generation (A3C3), which jointly optimizes neural network architectures and their hardware implementations to address the inefficiencies of traditional top-down AI system design flows. Conventional AI deployment often treats model design and hardware mapping as separate stages: an algorithm is first developed for accuracy, and only afterward adapted to meet latency, throughput, energy, or resource constraints. This separation can lead to suboptimal systems, particularly as modern AI workloads become increasingly heterogeneous, memory-intensive, and platform-dependent. A3C3 instead parameterizes both algorithmic and accelerator design spaces and searches them jointly, enabling the automatic generation of model-accelerator pairs that better balance accuracy, latency, throughput, energy efficiency, and hardware utilization. This article is a book chapter of the Handbook of Embedded Machine Learning, edited by Sudeep Pasricha and Muhammad Shafique, Springer Nature. --- ### Can LLMs Reason About Brand Ownership? An Empirical Study of Domain Attribution Intelligence Authors: Fathima Mashood, Mohamed Nabeel Published: 2026-06-18 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2606.20868v1 When a new domain resembling a popular brand appears, defenders face a fundamental ambiguity: it may be an attacker-created squatting site for phishing, or it may be a domain the brand itself registered, either defensively, to block attackers, or legitimately, for a new product or service launch. Incorrectly flagging a brand-owned domain as malicious produces a false positive that harms end users and damages the brand's reputation. Resolving this ambiguity requires brand intelligence: the ability to determine, at scale, whether a given domain belongs to a brand. Large language models (LLMs), with their broad knowledge of brand domain relationships, offer a promising zero configuration approach to this problem, but their reliability for brand intelligence tasks remains unknown. We present the first systematic empirical evaluation of LLM brand intelligence across three tasks: domain enumeration (Q1), open ended brand attribution (Q2), and binary ownership classification (Q3). We evaluate four models, Gemini 2.5 Flash, Gemini 3.5 Flash, Claude Sonnet 4.5, and Claude Sonnet 4.6, across four retrieval settings (in context, web search, WHOIS lookup, and combined) on 36 of the most phished brands. Our results reveal a stark dichotomy: models achieve up to 82% precision enumerating brand domains from memory alone, yet fail at ownership verification without external tools, with macro F1 at most 0.37 in ICL mode. WHOIS augmentation lifts Q3 macro F1 by up to 0.65 points, yielding near perfect precision (<= 0.99), dramatically reducing the false positive risk for defenders. We provide concrete recommendations for deploying LLMs in brand protection pipelines. --- ### FOCA: Future-Oriented Conditioning for Data-Efficient Vision-Language-Action Adaptation Authors: Duc Minh Nguyen, Nghiem Tuong Diep, Binh Gia Nguyen, Trong-Bao Ho, Doanh Le, Tan Q. Nguyen, Thien-Loc Ha, Nhiem Tran, Bao Thach, Nhat X. Tran, Tuan A. Tran, Artur Habuda et al. Published: 2026-06-18 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.20867v1 Vision-Language-Action (VLA) models enable general-purpose robotic control via large-scale multimodal pretraining, yet their effectiveness under few-shot imitation learning remains limited. We conduct a systematic stress test of state-of-the-art VLA models and show that performance degrades sharply as demonstrations are reduced, revealing a key weakness of existing adaptation strategies. To address this, we introduce FOCA, a future-oriented conditioning framework for data-efficient VLA adaptation. FOCA combines explicit prediction of task-grounded future interaction embeddings with implicit alignment to future goal observations, enabling long-horizon reasoning in latent space without pixel-level prediction. This formulation naturally supports action-free co-training with synthetic videos from video world models and can be interpreted as learning a future-conditioned value-like representation. Extensive experiments demonstrate FOCA achieves 95.7% success with 20 demonstrations on LIBERO, improves 7-12% on RoboCasa, and delivers up to 26% absolute gains on real robots, establishing a new state of the art in few-shot VLA adaptation. --- ### Synthetic Network Packet Generation through Statistical Learning and Genetic Algorithms Authors: Mayank Raj, Nathaniel D. Bastian, Lance Fiondella, Gokhan Kul Published: 2026-06-18 Categories: cs.CR, cs.LG arXiv: https://arxiv.org/abs/2606.20864v1 Developing robust intrusion detection systems (IDS) for IoT environments requires large, labeled datasets capturing realistic traffic distributions across both benign and malicious activity. Existing public datasets suffer from fixed activity distributions and extreme class imbalance, while deep generative models (GANs, VAEs) provide no mechanism to enforce that synthetic packets remain within physically valid feature ranges. This paper proposes and compares two constraint-enforcing approaches for synthetic IoT network packet generation: (i) a statistical learning method combining PCA-based latent space sampling with dual One-Class SVM (OCSVM) and Isolation Forest (IF) boundary enforcement, and (ii) a genetic algorithm (GA) method that treats packet generation as a multi-objective optimization problem with explicit fitness criteria for anomaly model acceptance and distributional fidelity. Both methods embed hard validity constraints -- dual anomaly-detection gating, feature-range clamping, and independent validation -- directly into the synthesis pipeline. Evaluation on the complete ACI IoT 2023 dataset (1,231,411 packets, 12 attack categories, class imbalance up to 175,805:1) demonstrates that both methods achieve PASS status across all categories under independently trained validators with a 30% anomaly rate threshold: the statistical method attains 1.20% average anomaly rate with ~1,091 packets/s throughput, while the GA attains 0.62% average anomaly rate with organic per-class variance (0.00%-2.50%) at ~5.7 packets/s. Both methods successfully amplify the 5-sample ARP Spoofing category by 200x to 1,000 validated packets. The ~190:1 throughput ratio between methods, combined with their complementary quality profiles, provides evidence-based selection criteria for deployment contexts ranging from rapid dataset augmentation to adversarial robustness testing. --- ### Betting on Moments: Legendre Jumper Martingales for Online Exchangeability Testing Authors: Johan Hallberg Szabadváry Published: 2026-06-18 Categories: stat.ML, cs.LG, math.ST arXiv: https://arxiv.org/abs/2606.20859v2 A fundamental assumption in statistics and machine learning is that ``the future looks like the past,'' formalized as exchangeability: the joint data distribution is order-invariant. In practice, this assumption is often violated due to distribution shifts over time. Early detection of exchangeability violations is crucial to prevent performance degradation and enable timely interventions like model retraining. Conformal test martingales offer a flexible, distribution-free framework for sequential exchangeability testing with guaranteed false-alarm rate control by betting against the uniformity of conformal p-values. While alternatives such as plug-in martingales and mixture-based strategies exist, computationally efficient baselines like the Simple Jumper are limited to detecting mean location shifts. We propose a family of conformal test martingales based on shifted Legendre polynomials that extend the Simple Jumper to higher-order moments. The Simple Legendre Jumper replaces linear betting functions with polynomials of arbitrary degree, enabling rapid detection of variance, skewness, and other higher-order deviations. The Product Legendre Jumper combines multiple polynomial degrees into a single betting function but suffers from exponential state-space growth, termed the jumping tax. To resolve this, we introduce the Variational Legendre Jumper, which employs a mean-field approximation to reduce complexity to constant time per step with minimal power loss, providing an expressive, scalable framework for real-time distribution shift monitoring. --- ### Evolutionary Discovery of Developmental Reward Schedules in Deep Reinforcement Learning Authors: Alan Nadelsticher Ruvalcaba Published: 2026-06-18 Categories: cs.LG, cs.NE arXiv: https://arxiv.org/abs/2606.20858v2 The temporal structure of reward composition in reinforcement learning (RL) is typically hand-designed and held fixed throughout training, leaving the progression of motivational priorities largely unexplored. In this work, we propose an evolutionary framework for discovering developmental reward schedules, in which three distinct biologically inspired motivational components -- agency, novelty, and reactivity -- are combined through time-varying weights that dynamically shift over the course of training. Evaluated on two sparse-reward MiniGrid tasks: DoorKey-6x6 and KeyCorridorS3R1, our framework compares the generalizability of four evolutionary algorithms: CMA-ES, xNES, DE, and L-SHADE against an extrinsically motivated baseline (our main comparison point), and three additional hand-designed methods. On DoorKey-6x6, all evolved methods outperform the non-evolved baselines, with L-SHADE achieving the best performance -- an approximate relative mean improvement of 11.4% over the extrinsic only baseline. On KeyCorridorS3R1, CMA-ES achieves the best overall performance, with the remaining evolved methods showing weaker and less reliable generalization capability compared to the extrinsic only baseline. Interestingly, the discovered schedules diverge from our defined developmental ordering, with novelty consistently emerging as the dominant early signal during training, across both tasks. Collectively, our results position evolutionary optimization as a promising approach for developmental reward schedule discovery in deep reinforcement learning, and suggest that what evolution finds to be optimal in computational settings may differ from what it finds to be optimal in biology. The code for this project can be found at: https://github.com/alannadels/Evolutionary_RL.git. --- ### SignVLA: Real-Time Sign Language-Guided Robotic Manipulation via Attention LSTM and Vision-Language-Action Models Authors: Ningwei Bai, Xinyu Tan, Harry Gardner, Zhengyang Zhong, Liuhaichen Yang, Luoyu Zhang, Zhekai Duan, Monkgogi Galeitsiwe, Zezhi Tang Published: 2026-06-18 Categories: cs.AI, cs.LG, cs.RO arXiv: https://arxiv.org/abs/2606.20857v1 Vision-Language-Action (VLA) models enable robots to execute manipulation tasks from natural-language instructions grounded in visual observations. However, existing VLA interfaces primarily rely on speech or text input, limiting accessibility for deaf, hard-of-hearing, and speech-impaired users. We present SignVLA, a real-time sign-language-guided VLA framework for accessible human-robot interaction. The system introduces a modular sign-to-text interface that converts visual sign gestures into semantic instructions compatible with downstream VLA policies. Given video streams, SignVLA extracts hand landmark features and employs an attention-enhanced Long Short-Term Memory (LSTM) network to capture temporal gesture dynamics for alphabet- and command-level sign recognition. A temporal stabilization module further improves prediction consistency in real-time interaction settings.The generated instruction sequence is then passed to a downstream VLA policy for sign-conditioned robotic manipulation. Experimental results demonstrate stable real-time sign recognition and successful execution of manipulation tasks driven by sign-language inputs. Our findings suggest that lightweight temporal sign recognition can serve as an effective and practical accessibility layer for multimodal embodied intelligence. --- ### Stochastic Signed Distance Processes Authors: Hiroki Sakuma, Masatoshi Okutomi Published: 2026-06-18 Categories: cs.CV, cs.GR, cs.LG arXiv: https://arxiv.org/abs/2606.20856v2 Multi-view surface reconstruction is a core problem in computer vision. One prominent line of work represents the surface implicitly as a signed distance field (SDF), optimizing it based on the photometric loss between rendered and observed pixel colors. These approaches typically employ SDF-based volume rendering to obtain a differentiable relaxation of discontinuous visibility along rays, thereby reducing reliance on silhouette supervision. In this paper, we reformulate SDF-based volume rendering as probabilistic surface rendering, where each pixel color is modeled as a mixture distribution induced by the random first ray-surface intersection. To this end, we introduce Stochastic Signed Distance Processes (SSDP), which model the SDF along each ray as a stochastic process, inducing a first-passage-time distribution for each ray. We then derive the first-passage probability for each sampling interval based on Bayesian filtering, together with its practical approximation for parallel rendering. We further show that NeuS, an existing SDF-based volume rendering method, arises as a special case of our formulation. Experiments on the DTU and MobileBrick datasets demonstrate that our method outperforms baselines in both surface reconstruction and uncertainty quantification, supporting the effectiveness of our first-passage formulation. Our code is available at https://github.com/skmhrk1209/SSDP. --- ### Translating Inference-Time Control to Radiology Vision-Language Models: Activation Steering for Pneumonia Classification on Chest X-rays Authors: Eduardo Moreno Judice de Mattos Farina, Mateus A. Esmeraldo, Felipe Akio Matsuoka, Paulo Eduardo de Aguiar Kuriki, Felipe Campos Kitamura Published: 2026-06-18 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.20852v1 Inference-time engineering can alter model behavior without fine-tuning. However, its utility for improving diagnostic performance in medical vision-language models (VLMs) remains unclear. We aim to evaluate whether Contrastive Activation Addition (CAA) can improve pneumonia classification in chest radiograph VLMs without updating model weights. Three frozen chest radiograph VLMs (MedGemma-4B-IT, NV-Reason-CXR-3B, and CheXOne-3B) were evaluated on the public Kermany pneumonia test set. Classification was based on the logits of the tokens Yes and No under a binary prompt. Steering vectors included a 30-pair answer-bias control, a 30-pair pneumonia text contrast, and an image-conditioned contrast derived from 30 pneumonia and 30 normal development images. A deterministic 200-image development set was used for layer and scale selection (100 images) and threshold calibration (100 images). Performance was assessed using ROC-AUC, PR-AUC, F1 score, threshold analyses, reverse-vector controls, random-vector controls, and conditional bootstrap confidence intervals. Fixed-threshold F1 improvements were frequently observed but did not consistently indicate improved diagnostic performance. For MedGemma-4B-IT. NV-Reason-CXR-3B showed the strongest benefit: calibrated F1 improved from 0.7692 in the zero-shot setting to 0.8619 with pneumonia-text steering and to 0.8727 with image-conditioned steering. For CheXOne-3B, pneumonia-text steering increased calibrated F1 from 0.8528 to 0.8666, although the confidence interval crossed zero. On this public pneumonia benchmark, CAA substantially altered prediction score distributions and operating characteristics without fine-tuning. Meaningful performance gains were observed in one of three evaluated VLMs, suggesting that activation steering may serve as a lightweight approach for adapting medical VLM behavior. --- ### From Uncertainty to Stability and Fidelity: Guiding Sparse-View 3D Gaussian Splatting with Fisher Information Authors: Junbao Zhou, Qingshan Xu, Yuan Zhou, Xiaolong Shen, Beier Zhu, Kesen Zhao, Yiming Zeng, Chen Bai, Cheng Lu, Hanwang Zhang Published: 2026-06-18 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.20842v1 3D Gaussian Splatting (3DGS) has emerged as a promising technique for novel view synthesis. However, 3DGS requires dense input views to achieve high-quality rendering. In sparse-view scenarios, 3DGS often prones to overfitting, resulting in noticeable artifacts and degraded rendering quality. Previous methods explore to address this issue by introducing additional priors (e.g. depth priors) or integrating regularization techniques (e.g. Dropout). However, these methods are often applied without principled guidance. In particular, prior-based augmentation typically samples novel viewpoints randomly, while Dropout-based regularization randomly removes Gaussians. The compounded randomness introduces uncertainty and instability, limiting the fidelity of novel view synthesis. In this paper, we propose a novel method for sparse-view 3DGS that incorporates Fisher Information to quantitatively guide the utilization of geometric priors and regularization. Specifically, our method comprises two key components: (1) Stereo augmentation with Fisher Information. By leveraging Fisher Information, we actively select most informative supporting views and use depth priors to curate reliable pseudo ground truths, which reduces randomness in augmentation and improves stability and rendering fidelity; (2) Uncertainty-aware regularization. We reduce the instability of Dropout-based regularization by using Fisher Information to quantitatively measure the uncertainty of each 3D Gaussian, and adaptively adjust the removal probability, leading to more stable and effective regularization. With these two components, our method effectively mitigates overfitting and improves the stability of optimization in sparse-view 3DGS, resulting in superior rendering fidelity. Extensive experiments show that our method achieves state-of-the-art performance in sparse-view novel view synthesis benchmarks. --- ### Process-Reward Tactic Evolution for Long-Horizon Bioinformatics Workflows Authors: Lingzhi Yang, Yubo Fan, Song Wu, Gilchan Park Published: 2026-06-18 Categories: cs.AI, cs.DC, cs.MA arXiv: https://arxiv.org/abs/2606.20839v1 LLM agents can write code and call tools, but reliable bioinformatics work requires long-horizon interaction with workflow software, typed data objects, provenance, and biological checks. We study this setting through Galaxy workflow execution. The agent must explore task data, construct or adapt an executable workflow DAG, bind inputs and dataset collections, monitor execution, debug failures, and validate biological outputs. We propose Process-Reward Tactic Evolution, a Galaxy-based training framework that turns verified workflow rollouts into reusable \tactics. During training, agents practice on curriculum-organized Galaxy tasks in Agent Gym; process verifiers score workflow construction, software interaction, execution, and biological correctness; successful and failed traces are distilled into a tactic library. At inference, the trained executor, Process-Reward Tactic Evolution, uses this library to execute held-out peer reviewed Galaxy workflow converted BioWorkflow Bench and BioAgent Bench tasks in isolated environments. The paper evaluates whether process-supervised tactic accumulation improves long-horizon bioinformatics workflow completion, biological correctness, and execution efficiency over no-memory and reflection-style baselines. --- ### Study on Quantitative Dynamic Epistemic Logic for Belief Revision Authors: Felipe Nunes de Souza Camargo Published: 2026-06-18 Categories: cs.AI, math.LO arXiv: https://arxiv.org/abs/2606.20837v1 Belief revision is a process in which an agent begins to believe in something she previously did not. I begin the paper by presenting, based on (Gärdenfors, 1998; Hansson, 1999), postulates for belief revision that constitute the basis of the AGM theory. I will then briefly show the semantics of a modal logic introduced in (van Ditmarsch, 2005), which I call `$P$'. This logic formalizes static epistemic states and has greater expressive power than AGM in doing so because it captures the quantitative notion of "degrees of conviction". The third step is to introduce revision operators on $P$ and, mostly following (van Ditmarsch, 2005), obtain the Dynamic Epistemic Logic (DEL) I call `$P*$'. It models processes of belief revision in several ways. Original results are presented in the following two sections. The first one of these sections revolves around a formalization of AGM postulates within $P*$ by proving some theorems related to the satisfaction of those postulates by revisions defined in $P*$. The last section features an analysis of $P*$'s revisions that go beyond the mere satisfaction of postulates. I compare their formal behavior with respect to some philosophical criteria. At last, I conclude that the functions presented in (van Ditmarsch, 2005) are not good formalizations of the philosophical intuition behind AGM. Instead, it is captured by the function $*^0$ originally defined in this paper (but highly inspired by (van Benthem, 2007)). An implementation of this function is also provided. --- ### ReLaTS: a Reinforcement Learning-based method for dynamically determining the coupling Time Step in multi-scale simulations of self-gravitating systems Authors: Veronica Saz Ulibarrena, Simon Portegies Zwart Published: 2026-06-18 Categories: astro-ph.IM, cs.LG arXiv: https://arxiv.org/abs/2606.20832v1 Astrophysical simulations frequently address multi-scale, multi-physics problems through subsystem decomposition, problem-tailored integration schemes, and coupling on fixed manually set timescales. Here we introduce ReLaTS, a reinforcement learning framework that dynamically selects the coupling time step to optimize the trade-off between accuracy and computational cost. We validate ReLaTS on star clusters containing a planetary system, and test the method by varying the number of stars $N_\star$ in the cluster and the number of planets ($N_{\rm planet}$) orbiting one of them. The method finds the optimal coupling time step that balances speed and accuracy without requiring expert knowledge. In addition, the trained network operates independently of the coupled \textit{N}-body algorithms, displaying stable performance across a range of setups. We observe that the method is less reliable for cases with infinitesimal masses, as their contribution to the total energy is negligible compared to that of the massive bodies, and the network is not capable of recognizing potential errors generated while integrating them. For long-time integration of large $N$ systems, the error accumulates. The reinforcement learning algorithm, however, manages to keep the energy error below a pre-set threshold. This approach substantially reduces energy errors relative to fixed-time step baselines without substantial additional computational overhead. Once trained, ReLaTS requires no expert tuning and generalizes across diverse astrophysical domains, enabling adaptive multi-scale simulations. --- ### NeoLoc-68: End-to-end 68-point neonatal facial landmark localisation in neonatal clinical environments Authors: Abdullah Bin-Obaid, Maria M. Cobo, Rebeccah Slater, Lionel Tarassenko, Mauricio Villarroel Published: 2026-06-18 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.20823v1 Facial landmark localisation is a prerequisite for developing automated, non-contact neonatal pain assessment methods. Clinicians use pain scales to judge the severity of pain, many of which rely on facial expression. However, facial landmark detectors trained on adult faces perform poorly in neonatal clinical environments due to frequent occlusions caused by medical equipment, varied head poses, and challenging imaging conditions, including motion blur triggered by sudden pain-related movements. We propose an end-to-end facial landmark detector capable of predicting 68 landmarks on neonatal faces in clinical environments. We combined 37,459 single-face images from 11 public datasets, standardised to 68-point markup, with 1,123 manually annotated frames from a neonatal research dataset (totalling over 76,000 landmarks). A YOLO-based keypoint model was adapted to regress the facial landmarks, initialised with weights from a pretrained neonatal face detector. On public datasets, our proposed model achieved state-of-the-art performance: Normalised Mean Error (NME) = 5.37, Failure Rate (FR) = 12.5%, Area Under the Cumulative Error Curve (AUC) at AUC0.08 = 38.00% and AUC0.1 = 48.70%. On the clinical neonatal test set, before fine-tuning, the model achieved the lowest Detection Failure Rate (DFR) = 5.3% among all baselines and showed strong generalisation. After fine-tuning, performance improved further to NME = 6.36, FR = 22.30%, DFR = 1.77%, AUC0.08 = 29.24% and AUC0.1 = 40.25%. To the best of our knowledge, this represents the first end-to-end 68-point neonatal facial landmark detection model. With further dataset expansion and refinement, it could support downstream tasks in neonatal health monitoring and pain-related facial analysis. --- ### CELEUS: Certifiable and Efficient LLM Evaluation via E-Processes Authors: Zhijian Zhou, Zesheng Ye, Zhaorun Chen, Bo Li, Feng Liu Published: 2026-06-18 Categories: cs.LG arXiv: https://arxiv.org/abs/2606.20820v2 Can we trust evaluation scores to capture an LLM's true real-world performance? Certifiable evaluation answers this question by providing guarantee for LLM evaluation. In particular, existing methods sequentially curate evaluation samples and keep updating confidence intervals (CIs) that cover the true performance with high probability (e.g., 95%) until some conditions are satisfied, e.g., the CI width reaches a target precision. However, existing methods are not generally anytime-valid: the claimed coverage (e.g., 95%) may fail when CIs are repeatedly updated and used to decide when to stop, leaving a gap between theoretical rigor and practice. This paper bridges this gap by proposing Celeus, a Certifiable framework for Efficient LLM evaluation, which leverages E-processes to build anytime-valid CIs. Concretely, we propose signals that combine two ingredients: (i) Uncertainty-guided sampling to select informative samples for evaluation, and (ii) Surrogate-assisted approximations for unevaluated samples. We prove that such signals remain unbiased for the evaluation score conditional on the past, enabling statistically-grounded and anytime-valid $e$-process CIs. More importantly, the two ingredients reduce estimation variance and help reach the target precision with fewer evaluated samples. We also prove that CIs obtained by Celeus can shrink at a near-parametric rate up to logarithmic factors and analyze the oracle variance-optimal sampling rule that motivates the empirical uncertainty-guided one. Experiments show that Celeus reaches the target precision using 54-62% fewer evaluated samples than baselines, while preserving anytime-valid coverage. --- ### Provably Sub-Linear Two-Timescale NeuroEvolution with Online Plasticity Authors: Shishen Lin, Yixin Chen Published: 2026-06-18 Categories: cs.NE, cs.LG arXiv: https://arxiv.org/abs/2606.20817v1 NeuroEvolution of Augmenting Topologies (NEAT) is a widely used neuroevolution algorithm for learning neural network architectures and weights for control tasks. However, standard offline optimisation searches for connection strengths directly, which can scale poorly in high-dimensional weight spaces and more difficult continuous control problems. Hybrid methods that combine neuroevolution with online learning can address this challenge, but their theoretical properties remain underexplored. This paper gives the first regret analysis for a general NeuroEvolutionary Online Learning (NEOL) framework, which decouples learning into two timescales: an outer loop for architecture search and an inner loop for online weight adaptation via rewardmodulated plasticity. Under mild conditions, we prove that NEOL achieves sublinear regret. Empirically, under fixed interaction budgets on four standard control benchmarks, a NEAT-based NEOL implementation achieves higher final fitness and lower variance than pure NEAT, and is competitive with strong reinforcement learning (RL) baselines on several tasks. The results are supported byWilcoxon rank-sum tests and ablation studies. Overall, the findings show that online plasticity can improve the sample efficiency and robustness of two-timescale neuroevolution. Code is available at https://github.com/boobaa2001/NeuroEvolution Online Learning NEOL. --- ### What Shapes Emergent Misalignment? Insights from Training Dynamics, Model Priors, and Data Authors: Yuchen Zhang, Anietta Weckauff, Diego Garcia-Olano, Maksym Andriushchenko Published: 2026-06-18 Categories: cs.AI, cs.LG, stat.ML arXiv: https://arxiv.org/abs/2606.20814v1 Emergent misalignment (EM) is a phenomenon in which models generalize with narrow fine-tuning, leading to broad (yet uneven) misalignment across evaluation questions. We study EM and its variability directly through the components of fine-tuning: training dynamics, model priors, and data. (1) We first explored how in-domain training loss relates to out-of-domain alignment scores across datasets and model families. Then, we tried to induce potential alternative local minima through different learning schedules for one narrow fine-tuning, but did not find strong runs with better broad alignment scores conditioned on similar or lower training loss. (2) We found that although the mean and standard deviations of the misaligned model scores are usually statistically different from those of the pre-trained model, there are some potential signals on overall positive correlation. The evaluation prompt-only activations from both the pre-trained and the original instruct models (prior to narrow fine-tuning) could predict fine-grained alignment scores after narrow fine-tuning. (3) Finally, we compared activation deltas before and after narrow fine-tuning and found moderate-to-high subspace overlap and similarity between the resulting activation shifts for training and evaluation prompts. Subspace overlaps between training and evaluation prompt activations correlate with their shifts' similarities when measuring with the last prompt-token activations. The train-evaluation data prompt overlap is controlled against overlap computed from random vectors and evaluation prompts activations. --- ### B(FM)$^2$: Brain Foundation Model via Flow Matching with SplitUNet Authors: Jaedong Hwang, Kathleen Zhang, Wei Dai, Konstantinos Kontras, Maarten Vanmarcke, Maarten De Vos, Ila Fiete, Paul Pu Liang Published: 2026-06-18 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2606.20812v1 EEG foundation models can learn generalizable representations from large-scale EEG corpora to enable single-backbone transfer across diverse clinical and brain-computer interface tasks. Existing models typically discretize the continuous multi-channel EEG waveform into patches or codebook tokens and train a transformer with masked self-supervision. Recognizing that this discretization fragments continuous brain rhythms and obscures fine-grained temporal dynamics, we present B(FM)$^2$(Brain Foundation Model via Flow Matching), whose inductive bias aligns with the data by pretraining directly on the raw signal using continuous-time flow matching without patches, tokenization, or masking. However, multi-channel EEG signals pose an architectural challenge for flow matching: time is densely sampled and highly autocorrelated (thousands of timepoints), while the electrode axis is short (tens of channels) at distinct scalp positions. To address this time-electrode asymmetry, we introduce SplitUNet, a velocity network that factorizes each block into separate 1D temporal and 1D electrode convolutions and downsamples only along time, preserving electrode topology throughout the hierarchy. B(FM)$^2$ sets a new state of the art on 7 of 9 standard downstream EEG classification tasks, using a pretraining budget of only 36,895 segments ($\approx$ 307h), 1-2 orders of magnitude ($\approx$ 30x) less than required by existing EEG foundation models. Further, it generates synthetic EEGs that two board-certified neurologists cannot distinguish from brain data (Cohen's $κ=$ -0.096). https://jd730.github.io/projects/BFM2 --- ### Eigenspace-Based Clustering for Personalized System Identification Authors: Abdulmoneam Ali, Dipankar Maity, Ahmed Arafa Published: 2026-06-18 Categories: eess.SY, cs.LG, eess.SP arXiv: https://arxiv.org/abs/2606.20811v1 We study the problem of system identification in heterogeneous settings, where different systems may follow distinct underlying dynamics. Existing clustered system identification approaches often rely on iterative training-based cluster assignment, which can be sensitive to learning uncertainty and model initialization. In contrast, we propose a one-shot, training-free clustering method that identifies similar systems using the structure of their locally observed data. Specifically, each system estimates a local state covariance matrix, and cluster identities are inferred by measuring the alignment between the leading covariance eigenspaces of different systems. We provide a mathematical interpretation of the proposed similarity score and develop a finite-sample analysis that characterizes how covariance estimation error induces eigenspace perturbations in terms of the underlying system dynamics. We then derive a probability bound for pairwise false merges and a global clustering success guarantee. Numerical experiments demonstrate that the proposed eigenspace-based clustering method effectively identifies systems with shared dynamics, leading to lower personalized model-estimation error compared with training-based clustering and non-clustered baselines. --- ### GroundShot: Visually Consistent Multi-Shot Long Video Generation via Entity-Grounded Shot Scheduling Authors: Yixuan Lai, Tianjia Shao, Weijia Dou, Siyu Zhu, Jingdong Wang Published: 2026-06-18 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.20799v3 Generating visually consistent multi-shot videos remains an open challenge. As videos span more shots, inconsistencies can accumulate across shots, causing entities that reappear across shots -- characters, objects, and locations -- to drift away from how they first appear. We observe that viewers judge consistency by comparing each later appearance of an entity with its first clear appearance; the visual quality of this initial appearance sets the consistency ceiling for all that follows. Motivated by this, we present \textbf{GroundShot}, a training-free, model-agnostic agentic framework for entity-grounded multi-shot generation. GroundShot builds an entity-level visual memory online from accepted generated shots: it schedules shots' generation order by their expected usefulness as entity references, grounds entities from generated videos, verifies their reliability before adding them to memory, and retrieves suitable entity references from memory before each shot is generated. To evaluate this entity-centered view of consistency, we further introduce \textbf{GroundBench}, a diagnostic benchmark that measures consistency at the entity level while isolating controlled challenge dimensions. Experiments show that GroundShot improves multi-shot consistency over existing methods while requiring no additional training or model modification. --- ### Fara-1.5: Scalable Learning Environments for Computer Use Agents Authors: Ahmed Awadallah, Sahil Gupta, Yash Lara, Yadong Lu, Hussein Mozannar, Akshay Nambi, Zach Nussbaum, Yash Pandya, Aravind Rajeswaran, Corby Rosset, Alexey Taymanov, Luiz do Valle et al. Published: 2026-06-18 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.20785v2 Collecting computer use data from human demonstrations is expensive and slow, motivating the need for scalable generation strategies. This requires two key ingredients: environments in which agents can act and verifiers that can judge whether their demonstrations succeeded. We introduce FaraGen1.5, a scalable data pipeline for computer use agents composed of three modular components: environments, solvers, and verifiers. FaraGen1.5 uses both live websites and synthetic environments that faithfully simulate domains gated by authentication or that require irreversible actions. It employs a solver harness that can be powered by multiple models, including strong frontier models such as GPT-5.4, and also incorporates a user simulator to enable multi-turn rollouts. Finally, FaraGen1.5 scores the resulting trajectories with three complementary verifiers covering task correctness, efficiency, and critical-point adherence. Using data produced by this pipeline, we train Fara1.5, a family of native computer use agents (CUAs) at three scales built on Qwen3.5 (4B, 9B, and 27B). To train these models, we employ a supervised finetuning (SFT) recipe that carefully balances data from FaraGen1.5 for broad coverage, specific high-value tasks, and target model deficiencies in an iterative approach. Each model sets a new state of the art (SoTA) for its size class on browser-use benchmarks: Fara1.5-9B reaches 63.4% on Online-Mind2Web and 86.6% on WebVoyager, while Fara1.5-27B achieves 72.3% on Online-Mind2Web, which is competitive with much larger proprietary systems. We also release weights for the Fara1.5 models under MIT license, making SoTA computer use accessible for all beyond closed API-only systems. --- ### World Action Models: A Survey Authors: Qiuhong Shen, Shihua Zhang, Yue Liao, Qi Li, Zhenxiong Tan, Shizun Wang, Shuicheng Yan, Xinchao Wang Published: 2026-06-18 Categories: cs.RO, cs.CV arXiv: https://arxiv.org/abs/2606.20781v1 World Action Models (WAMs) are embodied predictive-action models that make a forecast of the future available to action. Recent WAMs repurpose large video generation models, and a parallel line relies on language or vision-language backbones without a video-generation core. This rapid expansion has blurred the boundary among broad world models, video generation models, action-grounded video world models, Vision-Language-Action policies, and WAMs. This survey gives the field a common account. It first clarifies these boundaries, then organizes existing works through two complementary views. The first view asks what each method is required to generate, spanning rendered futures, latent futures, and video-generation-free action reasoning. The second view decomposes each method by predictive substrate, backbone, action coupling, and deployment regime. This anatomy supports a unified discussion of interactability, causality, persistence, physical plausibility, and generalization, followed by data, evaluation, and open challenges. Across these axes, a consistent design pattern emerges: WAMs are not simply video generators with action heads, but predictive-action methods whose design choices trade representational richness against compute, memory, latency, and action-label cost. The field is moving toward methods that generate less of the future while preserving what control requires. The survey homepage is available at https://world-action-models.github.io/. --- ### Formally Verified Code Synthesis for Structured Data Translation in a Medical Internet of Things Authors: Colin Samplawski, Adam D. Cobb Published: 2026-06-18 Categories: cs.SE, cs.AI, cs.NE arXiv: https://arxiv.org/abs/2606.20776v1 In this work we present a LLM powered, evolutionary code synthesis system for structured data translation in a Medical Internet of Things settings. A key challenge in this domain is ensuring that the synthesized code is trustworthy and reliable. To this end, we integrate a formal verification step into our code synthesis pipeline to ensure that any generated code is guaranteed to satisfy predefined requirements. In particular, we present a case study of integrating a novel device (a pulse oximeter) into the existing network of devices. Our system generates a formally verified translation between the device's JSON schema and the Fast Healthcare Interoperability Resources (FHIR) format used by the wider system. This formal verification stage ensures structured data translated by the generated code will always be in the target output schema. We provide a set of experimental results which demonstrate that our system is able to consistently generate correct translation at low cost. --- ### TriMotion: Modality-Agnostic Camera Control for Video Generation Authors: Seunghyun Shin, Jifei Song, Wooseok Jeon, Hae-Gon Jeon, Jiankang Deng Published: 2026-06-18 Categories: cs.CV, cs.AI, cs.RO arXiv: https://arxiv.org/abs/2606.20774v1 Camera motion control is essential for directing viewpoint changes in generative systems. However, existing methods typically condition the generation process on a single specific modality, such as explicit pose trajectories or reference videos, limiting their ability to support heterogeneous user inputs. To address this limitation, we present TriMotion, a modality-agnostic framework for camera-controlled video generation that maps video, pose, and text inputs, describing the same camera trajectory into a shared motion embedding space. Learning such a space requires synchronized supervision across modalities. Therefore, we build the Motion Triplet Dataset by extending a Multi-Cam Video Dataset with geometry-grounded motion descriptions derived from camera extrinsics. We further introduce a latent motion consistency objective that leverages the motion embedding space to encourage the generated video to follow the target camera trajectory directly in latent space, avoiding the cost of pixel-space decoding. Extensive experiments show that TriMotion generates high-quality videos that accurately follow the target camera trajectories across all three modalities. Beyond standard generation, the shared motion embedding space also enables flexible applications such as sequential motion composition and cross-modal motion interpolation. --- ### ELADO: Elliptic PDE Assessment Datasets for Operator Learning Authors: Frank Ehebrecht, Toni Scharle, Martin Atzmueller Published: 2026-06-18 Categories: cs.LG arXiv: https://arxiv.org/abs/2606.20771v1 We introduce ELADO (Elliptic PDE Assessment Datasets for Operator Learning), a systematic benchmark suite constructed to show and quantify failure modes of neural operator architectures when learning solution operators of elliptic PDEs. While the benchmarks of existing datasets focus on average case performance, the ELADO datasets are constructed to highlight challenges that arise naturally in elliptic PDE problems. In particular, we construct several datasets built around Poisson's equation and the Helmholtz equation, each with non-constant coefficients. We define a controllable data-generating process to create datasets, that are designed to isolate a distinct source of difficulty. Specifically, these are (1) heavy-tailed solution distributions arising from light-tailed coefficient field distributions, (2) spectral distribution shift of the input data, (3) heavy-tailed distributions in the frequency domain of solutions, arising from light-tailed coefficient field distributions, (4) input sensitivity of learned operators, quantified by an empirical local Lipschitz analysis, and (5) the effect of input signal complexity on prediction accuracy under controlled amplitude normalization. We evaluate several neural operator architectures across all datasets and show that heavy-tailed targets, spectral shift, and input sensitivity each cause substantial degradation of the prediction accuracy that standard datasets and metrics (e.g., the mean relative $L^2$ error) may obscure. --- ### Beyond 'One Language, One Script': Quantifying Orthographic Bias in Multilingual VLMs with PuMVR Authors: Prabhjot Singh, Bhushan Pawar, Madhu Reddiboina Published: 2026-06-18 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.20770v1 Current Vision-Language Models (VLMs) are celebrated for their multilingual capabilities, yet they operate under a flawed assumption: that one language corresponds to a single writing system. This overlooks billions of users of multi-script languages like Punjabi, Serbian, Hindi-Urdu, Kurdish, among many others, for whom a model's capability may be fractured by orthographic bias. We introduce PuMVR (Punjabi Multimodal Visual Reasoning), the first benchmark designed to quantify script-dependent bias through 375 culturally grounded image-reasoning tasks across Punjabi's three active scripts (Gurmukhi, Shahmukhi, Roman). Evaluating 10 state-of-the-art VLMs, we expose a substantial Script Gap: models frequently solve visual puzzles in one script while failing identical tasks in another, with accuracy deltas reaching 16% and Script Consistency Rates (SCR) as low as 24.8%. Crucially, visual input boosts absolute performance but does not close this gap, the relative bias persists. Our analysis suggests reasoning patterns show limited cross-script transferability, and Chain-of-Thought pathways diverge based on script alone. We propose SCR as a core metric for script-agnostic evaluation, challenging current multilingual assessment paradigms and providing a framework for equitable AI. --- ### FirstPass: Grounding AI Scientific Judgment in Multi-Round Editorial Outcomes Authors: Prabhjot Singh, Somnath Luitel, Manmeet Singh, Josh Durkee Published: 2026-06-18 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.20769v1 AI systems for peer review fail on three fronts: they train on Computer Science and Machine Learning venues alone, ignore the iterative dialogue that validates science, and evaluate on stylistic mimicry rather than real editorial judgment. We introduce FirstPass, a dataset and fine-tuned model that addresses all three. Curating 3,668 complete multi-round peer-review dialogues from Nature Communications across five scientific domains (biology, chemistry, neuroscience, physics, and earth science), we exploit mandatory transparent peer review (instituted November 2022) and verify 100% content integrity by automated audit. We fine-tune Qwen2.5-7B-Instruct via Low-Rank Adaptation (LoRA) on three tasks: review generation, reviewer updating, and revision-cycle prediction. Our key finding is that response-only loss masking is a prerequisite, not an optimization: without it, accuracy is 62.0%, below the majority baseline; with it, FirstPass achieves 80.5% accuracy and F1-macro 78.2% on predicting editorial outcomes (Standard vs. Extended revision cycles), outperforming Gemini-3.1-flash-lite-preview zero-shot by 10.4 percentage points and all baselines with statistical significance (McNemar p < 0.001). On generation, FirstPass produces reviews averaging 1,187 words, substantially closer to human references (2,155 words) than any baseline, achieving ROUGE-L 0.154 with significant gains over Qwen and DeepSeek zero-shot (p < 0.001). Deployed in the pre-submission loop as an anticipatory scientific co-author, FirstPass simulates expert critique and predicts revision cycle outcomes before submission, giving authors the judgment a trusted colleague would provide, with consistent cross-domain performance across five disciplines. --- ### UniSLAD: A Unified Framework for Structural and Logical Industrial Visual Anomaly Detection Authors: Changyi Li, Chao Yang, Yu Xiao, Kari Tammi Published: 2026-06-18 Categories: cs.CV, cs.AI, eess.IV arXiv: https://arxiv.org/abs/2606.20768v1 Visual anomaly detection is a fundamental task in industrial automation. While existing approaches have achieved notable progress in identifying structural defects, the detection of logical anomalies remains relatively underexplored. In practice, structural and logical anomalies frequently co-occur in industrial workflows. Therefore, a solution capable of detecting both structural and logical anomalies is crucial for advancing comprehensive anomaly detection research. To address this limitation, we propose a unified framework, termed UniSLAD, which jointly addresses logical and structural anomalies without additional training, enabling a practical solution for dynamic industrial environments. First, we introduce a dual-feature extractor that synergistically integrates a Convolutional Neural Network (CNN) backbone for local texture perception with a Transformer backbone for global contextual reasoning, yielding richer and more comprehensive representations. Building on this foundation, we design dual-granularity feature representation modules. At the patch level, memory banks enhanced by the Mahalanobis Transform (MT) preserve representative features and support more discriminative anomaly scoring. At the image level, distribution maps are aggregated using Lower-Upper Mean (LUM) and Power Mean Pooling (PMP), yielding a more robust global representation than conventional average pooling. Extensive experiments on the two industrial benchmarks demonstrate that UniSLAD achieves competitive performance in comprehensive anomaly detection, achieving 99.4% and 93.1%, respectively. Furthermore, ablation studies verify the individual contributions and effectiveness of each proposed component. --- ### Dataset-Aware Cold-Start Active Learning for Annotation-Efficient 3D Medical Image Segmentation Authors: Rémi Hattat, Marine Beaumont, Charline Bertholdt, Gabriela Hossu, Olivier Morel, Bailiang Chen Published: 2026-06-18 Categories: eess.IV, cs.LG, q-bio.QM arXiv: https://arxiv.org/abs/2606.20765v1 Deep learning for 3D medical image segmentation requires extensive manual annotations, a major bottleneck in volumetric medical imaging. Active learning aims to reduce this burden by selecting informative samples for annotation, but most methods assume that an initial labeled set is already available. This leaves the cold-start problem largely unresolved: how to select the first volumes from a fully unlabeled pool before any task-specific model is trained. We propose CSCS, a Curriculum-Stratified Cold-Start framework that adapts initial sample selection to the structure of the unlabeled dataset. CSCS combines two self-supervised, label-free signals: local typicality, measuring representativeness in the embedding space, and reconstruction-based uncertainty, used as a proxy for sample difficulty. These signals are combined through a weighted geometric score, where the weighting is determined by a closed-form pacing rule based on the effective annotation budget and the Difficulty-Coverage Ratio, a pool-level statistic measuring the alignment between difficulty and representativeness. We evaluate CSCS on four 3D medical image segmentation benchmarks: BraTS, FeTA, Spleen, and an in-house fetal MRI dataset. Using nnU-Net as downstream segmentation model, CSCS shows consistently competitive performance across datasets and annotation budgets, with the strongest gains in low-to-mid annotation regimes. These results suggest that dataset-aware cold-start initialization can improve the robustness of active learning for 3D medical image segmentation by adapting sample selection to the geometry of the unlabeled pool. --- ### One Image is All You Need: Agentic One-Shot Image Generation via Text-Based World Models for Long-Tail Spatial Perception Authors: Keqin Zeng, Shuting Su, Shihao Lin, Ziyue Li, Rui Zhao Published: 2026-06-18 Categories: cs.CV, cs.AI, cs.GR, cs.LG arXiv: https://arxiv.org/abs/2606.20764v1 Reliable spatial decision automation, such as autonomous driving and maritime surveillance, critically depends on robust visual perception. However, real-world spatiotemporal data exhibits severe heterogeneity, often manifesting as extreme long-tail distributions for safety-critical scenarios. This data scarcity induces dataset shift that degrades detection performance and pose safety risks. While synthetic data generation offers a potential solution, existing generative approaches, such as diffusion models and Generative Adversarial Networks (GANs), often lack explicit spatial grounding and structural constraints, resulting in spatial and physical inconsistencies in generated scenes. To address these challenges, we introduce WMGen-v1, an agentic text-based world model framework for long-tail spatial data generation. WMGen-v1 employs a Large Vision-Language Model (LVLM) to construct a structured scene representation from a single reference image, while a Large Language Model (LLM) performs guidance-based scene expansion under physical plausibility and commonsense constraints. Subsequently, conditioned on the structured semantic representations produced by this reasoning process, a diffusion model generates diverse and physically grounded long-tail training data. Experiments on internal industrial datasets, ROADWork, and LaRS benchmarks demonstrate that WMGen-v1 outperforms baseline approaches. Notably, detectors trained solely on WMGen-v1 synthetic data approach real-only performance on aggregate dataset-level metrics, highlighting its potential to alleviate long-tail data scarcity for downstream spatial perception. --- ### Co-Construction Blindness and Asymmetric Epistemic Vulnerability in Human-LLM Interaction Authors: Bianca Helena Ximenes Published: 2026-06-18 Categories: cs.HC, cs.AI, cs.CY arXiv: https://arxiv.org/abs/2606.20762v1 This paper introduces two constructs to describe, as far as we know, a previously unnamed risk in human-LLM interaction. Co-construction blindness is the failure to recognize that LLM outputs are not independent assessments to be verified, but co-constructed artifacts shaped by the user's own inputs, accumulated history, and metadata. Every user of a conversational LLM is IN the loop, not ON it -- yet every deployment disclaimer positions them as external auditors. Asymmetric epistemic vulnerability describes the condition in which co-construction blindness produces consequences of radically different magnitude depending on where in the authority structure the user sits. We argue that these constructs describe a structural inevitability, not an anomaly, using the public case of Richard Dawkins's interaction with Claude as a paradigmatic instance. We document a secondary mechanism -- structural deference -- through a first-person exchange in which a large language model concedes that it treated Dawkins more gently than warranted because his intellectual output is represented in its training data. We map the research gaps this analysis opens and call for shared terminology as a precondition for appropriate governance and design response. --- ### Integrating Large Language Model Agents with Digital Twins for Industrial Autonomous Systems Authors: Yuchen Xia Published: 2026-06-18 Categories: cs.SE, cs.AI, cs.ET, cs.MA, eess.SY arXiv: https://arxiv.org/abs/2606.20761v1 Industrial automation is being transformed by digitalization and the increasing use of cyber-physical systems. Modern production environments require greater adaptability, faster reconfiguration, and more intuitive human-machine interaction. However, traditional rule-based systems rely on fixed logic and cannot autonomously adapt to changing conditions. Consequently, current automation systems lack a systematic approach for integrating adaptive and generalizable reasoning capabilities for interpreting, planning, and executing user tasks across dynamic environments and heterogeneous components. This dissertation proposes a three-layer framework that integrates large language models (LLMs), digital twins, and automation systems into an autonomous system. Autonomy is defined as a design property assigned to system components and enabled through LLM-based reasoning to achieve adaptive, goal-oriented behavior. The Task-Process-Service-Resource (TPSR) model is introduced to transform user tasks into executable processes. Four LLM roles are identified: process orchestration, service matching, digital resource generation, and agent-as-a-service. Five peer-reviewed studies develop and refine these concepts using the design science research methodology. Case studies and prototypes demonstrate adaptive task planning, event-driven control, simulation-based parameterization, and digital model generation. Results show high task executability, command correctness, and content-generation accuracy while reducing manual effort. The framework enables the integration of LLM-based reasoning into industrial automation systems and improves adaptability and usability. Limitations include dependence on accurate digital representations, the computational demands of LLMs, and the need for human intervention in safety-critical situations. --- ### A Topology-Aware, Memory-Centric Architecture that Separates Root-Cause Derivation from Root-Cause Explanation Authors: Momil Seedat Published: 2026-06-18 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2606.20758v1 Modern microservice deployments fail in ways that are easy to detect and hard to explain. When a fault propagates along service dependencies, alerts fire in floods, dashboards multiply, and the scarce resource, an engineer who understands how the services relate, is consumed reconstructing context that the monitoring stack discarded. We argue that the missing ingredient in autonomous operations is not a better anomaly detector or a larger language model, but operational memory: a persistent, structured representation of how a system normally behaves, how its parts depend on one another, and how it has failed before. We present O PS C ORTEX, a working multi-agent prototype that organizes this memory into four tiers and uses it to separate two tasks the field usually conflates: deriving a root cause and explaining it. Root cause is computed deterministically from a learned dependency graph and the temporal ordering of threshold crossings; a large language model (LLM) is then asked only to explain, confirm, and recommend, using evidence the system has already assembled. We motivate the design with two documented production cascading failures, review representative literature on observability, anomaly detection, graph-based localization, and LLM-assisted diagnosis, and show how each architectural choice maps directly to a failure mode those incidents exhibit. The prototype is validated on an instrumented e-commerce benchmark with eight injectable failure scenarios. --- ### Evidential Fusion Network for Multimodal Survival Prediction under Missing Modalities Authors: Yucheng Xing, Hailan Mo, Zi Wang, Ling Huang, Mengling Feng Published: 2026-06-18 Categories: cs.LG arXiv: https://arxiv.org/abs/2606.20757v2 Recent multimodal survival prediction models have demonstrated strong predictive performance by leveraging complementary information across modalities. However, such models generally assume data completeness and exhibit limited robustness toward missing modalities, which are frequently encountered in real-world clinical settings. We propose the Evidential Missing Modality Survival Fusion (EMMS) model for multimodal survival prediction under missing modalities. EMMS offers a straightforward, computationally effective approach to survival analysis without requiring a generative phase for missing data. By employing Dempster-Shafer theory and Gaussian Random Fuzzy Numbers for multimodal decision fusion, it considers both aleatoric and epistemic uncertainty alongside modality reliability for fusion. Moreover, the model treats missing modalities as vacuous evidence, preventing interference with available inputs and naturally reflecting increased uncertainty and calibrated predictions. Extensive experiments on four cancer datasets demonstrate state-of-the-art performance while providing calibrated and interpretable uncertainty estimates under incomplete multimodal observations, without introducing additional computational overhead. --- ### Evidential Fusion Network for Multimodal Survival Prediction under Missing Modalities Authors: Yucheng Xing, Hailan Mo, Zi Wang, Ling Huang, Mengling Feng Published: 2026-06-18 Categories: cs.LG arXiv: https://arxiv.org/abs/2606.20757v1 Recent multimodal survival prediction models have demonstrated strong predictive performance by leveraging complementary information across modalities. However, such models generally assume data completeness and exhibit limited robustness toward missing modalities, which are frequently encountered in real-world clinical settings. We propose the Evidential Missing Modality Survival Fusion (EMMS) model for multimodal survival prediction under missing modalities. EMMS offers a straightforward, computationally effective approach to survival analysis without requiring a generative phase for missing data. By employing Dempster-Shafer theory and Gaussian Random Fuzzy Numbers for multimodal decision fusion, it considers both aleatoric and epistemic uncertainty alongside modality reliability for fusion. Moreover, the model treats missing modalities as vacuous evidence, preventing interference with available inputs and naturally reflecting increased uncertainty and calibrated predictions. Extensive experiments on four cancer datasets demonstrate state-of-the-art performance while providing calibrated and interpretable uncertainty estimates under incomplete multimodal observations, without introducing additional computational overhead. --- ### A large-scale foundation model enables simulation-to-real adaptation for nuclear magnetic resonance-based molecular structure analysis Authors: Chen Yang, Zheng Fang, Hanyu Sun, Fanjie Xu, Hongxin Xiang, Hanyu Gao, Xiangxiang Zeng, Yuqiang Li, Xiaojian Wang, Jun Xia Published: 2026-06-18 Categories: physics.chem-ph, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.20756v1 Nuclear Magnetic Resonance (NMR) spectroscopy is a powerful tool for molecular structure analysis, and spectral artificial intelligence offers great potential for its rapid and automated interpretation. However, the scarcity of experimental NMR datasets has constrained deep learning in this domain to narrow, task-specific applications that lack broad generalization. Here, we introduce UltraNMR, a large-scale foundation model for NMR that leverages the intrinsic properties of NMR spectra to learn generalizable spectral representations. We collected 158 million paired simulated $^{1}$H and $^{13}$C NMR spectra to train UltraNMR, employing multiple domain-specific pre-training objectives. UltraNMR captures both intra-spectral and inter-spectral dependencies, enabling seamless simulation-to-real adaptation. We demonstrate that adapting UltraNMR to a range of molecular structure analysis tasks on experimental NMR spectra consistently yields state-of-the-art performance and clearly outperforms UltraNMR variants trained directly on downstream data without simulation pre-training. We also construct a large-scale NMR spectral vector library by encoding simulated NMR spectra using UltraNMR, covering 94 million unique molecules and enabling effective structure-aware retrieval. In real-world applications, UltraNMR facilitates the structural elucidation of two previously unknown natural products from Chinese herbal medicines recorded in the Chinese Pharmacopoeia. These results suggest that large-scale simulation pre-training can effectively bridge the simulation-to-real gap, enabling robust and generalizable molecular structure analysis of real-world NMR spectra. ---