# Research Papers — 2026-06 · bundle page 104 of 260 > Full-text markdown of papers #5151–5200 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/104 [← Previous bundle](https://aipod.dev/papers/archive/2026-06/page/103.md) [Next bundle →](https://aipod.dev/papers/archive/2026-06/page/105.md) ### ELEVATE: Designing Human-Centered GenAI Virtual Tutors for Scalable and Inclusive Education Authors: Lorenzo Stacchio, Michele Giordano, Daniele Berardini, Primo Zingaretti, Emanuele Frontoni Published: 2026-06-17 Categories: cs.CY, cs.AI arXiv: https://arxiv.org/abs/2606.30662v1 The advent of Generative Artificial Intelligence (GenAI), and in particular Large Language Models (LLMs), is reshaping educational practice, while intensifying ethical debate about its adoption. To date, the dominant paradigm remains cloud-based and text-only chatbot: a centralized service that offers limited pedagogical control, weak transparency over knowledge sources, and non-trivial risks for privacy and regulatory compliance. This model also presumes continuous connectivity and recurring API costs, creating structural barriers for many institutions, reinforcing existing digital divides. At the same time, educational interaction with LLM can benefit from multimodal cues and embodied presence, requiring interfaces that move beyond text-only tutoring. In this work, we propose ELEVATE (Efficient LLM Education with Virtual Avatar Teaching Engine), a framework to develop efficient GenAI-driven avatar tutors governed by epistemic infrastructures. ELEVATE integrates LLM-driven dialogue with embodied 3D avatars for multimodal interaction and adopts a local-first execution model enabling deployment on consumer-grade hardware. The framework formalizes a three-stratum design that separates (i) a student-facing virtual avatar interaction layer, (ii) a local GenAI execution and multimodal synthesis core, and (iii) a teacher-facing governance layer. We implemented and evaluated a working prototype deployed in a real-world educational curriculum. The system runs on standard PCs and smartphones, and we provide system-level performance evidence to show responsive interaction under realistic hardware constraints. Finally, we discuss sociotechnical and pedagogical implications for responsible adoption, positioning ELEVATE as a scalable pathway for privacy-preserving and inclusive GenAI tutoring across heterogeneous school environments. --- ### Recursive Self-Evolving Agents via Held-Out Selection Authors: Michael Nguyen, Quoc Nguyen, Paul Vuong Published: 2026-06-17 Categories: cs.AI arXiv: https://arxiv.org/abs/2606.28374v1 LLM agents are increasingly improved without weight updates by evolving a natural-language artifact, such as reflections, workflows, playbooks, cheatsheets, or optimized prompts, that conditions a frozen policy. Such methods are typically reported as wins on the single benchmark where they help. We study them apples-to-apples and surface a sharper picture. We introduce RSEA, a Recursive Self-Evolving Agent that carries a compact three-layer natural-language state: an imperative strategy, reusable skills, and a procedural playbook. Across generations, RSEA rewrites all three layers from its own trajectories and commits a candidate only if it does not regress on a disjoint held-out split, using a strict keep-better gate. Across four diverse benchmarks, ALFWorld, GAIA, (τ)-bench, and WebShop, and six faithful baselines, ReAct, Reflexion, GEPA, AWM, ACE, and Dynamic Cheatsheet, all evaluated on one shared local backbone, we find three main results. First, no artifact universally wins. RSEA is the strongest single-pass method on ALFWorld, reaching 69.3% compared with 64.6% for ReAct (McNemar (p=0.015)), and reaches 79.4% with retry, the best overall result. However, concrete-workflow induction, represented by AWM, is best on the strong-backbone tool-use tasks. Second, unguarded context evolution is high-variance and unsafe. Dynamic Cheatsheet, which curates context online without a held-out gate, is near-best on ALFWorld at 70.7%, yet collapses on WebShop, with a score of 0.14 compared with 0.43 for ReAct. Third, RSEA's strict held-out selection is what makes recursive self-evolution monotone-safe: it never significantly underperforms the base agent on any benchmark and falls back to vanilla ReAct when evolved context would hurt. --- ### Model Merging to Evolution: Parameter Space Exploration for Expert Models Authors: Chao Wang, Yuchen Guo, Zheng Tan, Guanchun Wang, Yanbiao Ma, Qiqi Duan, Peng Wu Published: 2026-06-17 Categories: cs.NE, cs.AI arXiv: https://arxiv.org/abs/2606.28373v1 Model merging integrates the capabilities of multiple expert models to create strong models for multiple tasks without additional training, thereby reducing computational resource requirements. However, existing methods operate within the convex combination space of expert models, failing to explore high-performance regions outside this space. This paper proposes the MERGEvolve framework, which unifies model merging and evolution within an evolution strategy by treating the merged model as the initialization for evolutionary exploration of the parameter space. During the merging phase, expert models act as deterministic sources to build a strong initial point. The evolution phase then explores the parameter space using random noise. Theoretical analysis shows that MERGEvolve explores regions outside the convex combination space. Extensive experiments on single-task and multi-task benchmarks demonstrate that MERGEvolve consistently achieves performance competitive with advanced model merging baselines. Ablation studies confirm that a high-quality initial point is critical for efficient exploration of the parameter space. --- ### Constraint-aware Optimization in Auto-Tuning Authors: Floris-Jan Willemsen, Stijn Heldens, Rob V. van Nieuwpoort, Ben van Werkhoven Published: 2026-06-17 Categories: cs.NE arXiv: https://arxiv.org/abs/2606.28372v1 Automatic performance tuning, or auto-tuning, is a key technique in high-performance computing, enabling applications to adapt to complex and evolving hardware architectures. A central challenge is the need to optimize over large discrete, constrained parameter spaces, where many candidate configurations are invalid due to hardware or software correctness constraints. Traditional evolutionary algorithms, such as Differential Evolution, Particle Swarm Optimization, and Genetic Algorithms, are not inherently constraint-aware and thus often waste computational resources evaluating invalid solutions. In this work, we present and evaluate constraint-aware variants of four evolutionary algorithms for auto-tuning. Through extensive experiments on a representative benchmark suite, we show that constraint-aware optimization leads to faster convergence and improved performance over unconstrained methods. Furthermore, we demonstrate that our methods outperform the pyATF methods, a state-of-the-art framework for constraint-based auto-tuning. Our results demonstrate that incorporating constraint-awareness into the optimization process significantly enhances their applicability and effectiveness in real-world auto-tuning problems. Constraint-awareness improved algorithm efficiency by ~39 on average, correlated with search space sparsity. The algorithms developed in this study are publicly available as open-source contributions to the Kernel Tuner framework, facilitating future research and benefitting users. --- ### GeoISF: Instance Semantic Forest Inspired Large-Scale Cross-View Geo-Localization via Ground LiDAR-to-Satellite Image Authors: Di Hu, Xia Yuan, Chunxia Zhao Published: 2026-06-17 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.28371v1 The problem of localization on a large-scale satellite image given a frame of query ground view point clouds remains challenging. Existing LiDAR-to-image cross-view localization methods struggle in large-scale scenarios due to limited semantic alignment and the modality gap between point clouds and satellite images. This paper introduces the large-scale LiDAR-to-image geo-localization pipeline called GeoISF. GeoISF introduces an instance semantic forest constructed using WordNet, which enhances temporal semantic representation and discriminative power by integrating semantic trees from multiple frames. By leveraging environmental semantic representation as a shared medium, GeoISF effectively bridges the modality gap and improves semantic matching accuracy. Extensive experiments demonstrate the superior performance of GeoISF in large-scale cross-view localization, 13.22 times better than the parallel LiDAR-to-image method in the R@10 metric on the KITTI dataset. The proposed method addresses the existing gap in large-scale LiDAR-to-image cross-view localization, offering a robust solution to the computational and accuracy challenges inherent in such scenarios. We will release the code as an open-source resource available online for the broader research community. --- ### A Hybrid Quantum-Classical Approach for Melt Pool Prediction in Laser Powder Bed Fusion Authors: Matthew M. Sato, Kincho H. Law Published: 2026-06-17 Categories: quant-ph, cs.LG arXiv: https://arxiv.org/abs/2606.23719v1 Laser powder bed fusion (LPBF) is a promising additive manufacturing technique that suffers from quality assurance concerns. Predicting melt pools from process parameters is crucial for assessing quality prior to manufacturing but remains a difficult problem because of the complex physical processes underlying LPBF. Quantum computers present a new computing paradigm, providing a new approach to information processing using quantum entanglement and superposition. This paper presents a practical demonstration of a hybrid quantum-classical model that leverages quantum computing to improve process parameter feature extraction with a quantum feature encoder. To make the quantum approach computationally feasible for large datasets, we first employ a clustering algorithm to reduce the number of expensive quantum computations. These quantum features are then processed by a classical neural network to predict the melt pool morphology, allowing for more accurate predictions of melt pools. We demonstrate the method using a quantum simulator, analyze the effect of measurement shot noise on the predictive performance of the network, and verify the results using quantum hardware. Finally, by examining which quantum features are most important, we provide insights that can inform the future design of more effective quantum encoding circuits. Ultimately, the performance improvement over purely classical networks validates the hybrid approach, demonstrating an engineering application of quantum computing using noisy and intermediate scale quantum (NISQ) devices. --- ### Dimensionality Reduction of QAOA Parameter Space with Kernel PCA for Max-Cut Authors: Sidharth Brahmandam, Vayd Ramkumar Published: 2026-06-17 Categories: quant-ph, cs.LG arXiv: https://arxiv.org/abs/2606.23718v1 The Quantum Approximate Optimization Algorithm (QAOA) is a leading variational algorithm for combinatorial optimization on near term quantum devices. As circuit depth increases, the number of optimization parameters grows, making the search landscape increasingly nonlinear and difficult to optimize. Previous studies have shown that optimal QAOA parameters often lie on a low dimensional manifold that can be approximated using Principal Component Analysis (PCA) at shallow circuit depths. However, the effectiveness of PCA decreases at higher depths because the underlying parameter manifold becomes increasingly nonlinear. In this work, we investigate Kernel Principal Component Analysis (KPCA) with a radial basis function kernel as a nonlinear dimensionality reduction technique for QAOA parameter optimization. The model is trained using 200 graphs from each of 3 graph families, namely Erdos-Renyi, Barabasi-Albert, and Watts-Strogatz, with graph sizes ranging from 7 to 10 nodes. Performance is evaluated on 30 test graphs containing 12 nodes at circuit depths 1, 2, 4, and 8. Experimental results demonstrate that KPCA consistently outperforms PCA at deeper circuit depths across all graph families. At depth 8, KPCA achieves approximation ratios above 0.86, while PCA declines to approximately 0.81 to 0.83. Both methods reduce the number of quantum circuit evaluations by more than 93 percent relative to unrestricted QAOA optimization. These findings suggest that nonlinear kernel methods more effectively capture the structure of the QAOA parameter manifold and provide a practical approach for scaling variational quantum optimization to deeper circuits. --- ### Self-Improvement Can Self-Regress: The Rise-and-Collapse Failure Mode of LLM Self-Training Authors: Jianzhe Lin Published: 2026-06-17 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.21090v1 Self-improvement can self-regress. In REINFORCE post-training for code, a model can quickly improve on its optimized metric and then collapse within the same training campaign. We study this in a controlled multi-seed testbed using Qwen-2.5-3B and Qwen-2.5-7B, trained on competitive-programming tasks with binary CodeGrader reward across 10 sequential 20-step campaigns. Across campaigns, pass@1 shows a robust rise-then-collapse pattern: it peaks within tens of gradient steps and then falls back, sometimes to near zero. This is not cross-task catastrophic forgetting, but within-task policy over-optimization on a fixed distribution; KL- and EWC-style constraints do not prevent it. We ask where the control loop should sit. We compare three levels: CARE, a between-campaign memory mechanism with a capability posterior, transfer gate, and regression-aware belief revision; ES, a within-campaign early-stop rule that rolls forward the peak checkpoint and sets the next budget to peak_step+3; and GRPO, which changes the RL update using group-relative reward normalization. The answer is regime-dependent. On Qwen-2.5-3B, where naive REINFORCE is fragile, CARE v2 nearly doubles end-of-chain pass@1 from 4.9% to 9.5%, with paired bootstrap 95% CI (+0.4,+8.9) and gains in 4/5 seeds. On Qwen-2.5-7B, CARE reaches parity with naive REINFORCE, 13.8% vs. 11.8%, while ES reaches 22.2% (14.1,28.0). Out-of-the-box GRPO reaches 20.7% (15.7,25.1), nearly matching REINFORCE+ES. GRPO raises the floor but does not remove the cliff. Its 7B gain mainly comes from better between-campaign carryover, while the within-campaign peak-to-end gap remains about 17 points under both REINFORCE and GRPO. GRPO+ES gives mixed evidence: 2/3 seeds improve, but one final cliff lowers the mean to 17.0% (0.0,28.1). A Gemma-3-4B pilot shows the same signature, suggesting the phenomenon is not limited to Qwen. --- ### Repeated post-training is not Self-improving: Diagnosing Scientific Amnesia in Continual DPO Pipelines Authors: Jianzhe Lin, Fei Wang, Xiaolin Li, Rajeshkumar Golani, Jubin Chheda Published: 2026-06-17 Categories: cs.AI arXiv: https://arxiv.org/abs/2606.21089v1 Industrial LLM teams often ship behavior updates by repeatedly DPO-training a base model on sequences of related preference-data campaigns. The dominant failure mode in this regime is not always classical catastrophic forgetting: a pipeline may preserve previously learned behaviors while still failing to accumulate reusable methodological knowledge about how to train the next campaign. We call this failure mode scientific amnesia. This paper turns that practitioner intuition into a measurable industrial problem. We contribute: (i) a diagnostic suite for amnesia, (ii) a Program-based pipeline that chains FSDP-sharded DPO checkpoints across Qwen2.5-7B-Instruct runs, (iii) a 30-campaign HumanEval subdomain benchmark, and (iv) a comparative diagnostic study of five strategy proposers: random memory, rule-based scheduling, retrieval-only memory, warm-start Bayesian optimization, and MSCL, a meta-scientific memory and reasoner candidate. Across a single-seed 5-condition * 3-step real-LM chain, 4 of 5 candidates degrade in step-level peak pass@1, including MSCL; only the deliberately conservative rule-based schedule improves. Follow-up pilots qualify rather than overturn this finding: in a heterogeneous chain, MSCL is the only completed candidate that improves, whereas in a small multi-seed homogeneous sweep, retrieval-only has the best mean Delta and no pairwise candidate gap is statistically distinguishable. The contribution is therefore diagnostic, not a claim that MSCL solves the problem: scientific amnesia is observable in a production-like continual-DPO pipeline, and conclusions about interventions depend sharply on chain regime, evaluator design, and seed coverage. --- ### CIExplainer++: Generating Causal and Interpretable Explanations for Graph Neural Networks Authors: Francisco Caldas, Sahil Satish Kumar, Ruben Belo, Cláudia Soares Published: 2026-06-17 Categories: cs.LG, stat.OT arXiv: https://arxiv.org/abs/2606.20747v1 Explainable Artificial Intelligence aims to make black-box models more trustworthy by presenting, in a human-understandable manner, the elements that lead to the model's output. This involves both (i) identifying components and connections with genuine causal influence on outputs and (ii) translating such structures into an interpretable representation. For the former, we introduce CIExplainer, a novel perturbation-based method grounded in causal inference for explaining Graph Neural Networks (GNNs). CIExplainer identifies the subgraph with the highest causal effects on GNN predictions using the Potential Outcome Framework. We evaluate and compare CIExplainer on various GNN architectures (GCN, GraphSAGE, GAT, GIN) and datasets. To bridge subgraph explanations with human interpretability, we further propose G2TeXplainer, a method that transforms causal subgraphs into natural language explanations that capture both feature-level and relational information. --- ### Massive Activations Are Architecturally Robust: A Controlled Scratch/Commitment Residual Stream Test Authors: Maruthi Vemula Published: 2026-06-17 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2606.20743v1 Trained transformers reliably develop massive activations, a small number of hidden dimensions whose magnitude is far above the median and which concentrate on the sequence-start token. Whether these outliers are a removable artifact of the residual stream's overloaded read and write role, or instead a functional necessity, is actively debated. We test the artifact hypothesis directly, with an architectural intervention. Our architecture, Ledger Residuals, splits the residual stream into a mutable scratch stream (Deliberation) that intermediate computation may freely overwrite and a protected, decode-only accumulator (Commitment) that holds the representation the model reads out. If massive activations exist only because one stream is forced to be both scratchpad and answer, then a dedicated answer channel should remove the need for them. We find that it does not. In matched-loss language models at the 160M and 290M scales, the model rebuilds the canonical fixed-dimension, start-token outlier inside the protected channel. The rebuilt feature is smaller in magnitude than in a standard transformer but more sharply concentrated on the start token, and a stronger sparsity penalty makes it more persistent and more concentrated still, rather than removing it. Massive activations therefore look architecturally robust: they re-emerge in whichever representation the model decodes from, which is what we would expect if they are functional rather than incidental. We release our architecture and measurement code. --- ### A Digital Twin Framework for Traffic-Aware UAV Pavement Monitoring in Open-Traffic Conditions Authors: Yamil Uchani, Grace Luna, Edwin Salcedo, Mauricio Figueroa Published: 2026-06-17 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2606.20742v2 UAV-based pavement inspection can reduce the cost and risk of road-surface monitoring, but real-world deployment remains difficult when traffic, pedestrians, and temporary occlusions affect defect visibility. This paper presents a Unity-based digital twin framework for traffic-aware UAV pavement monitoring in open-traffic conditions. The proposed environment integrates procedurally generated road defects, dynamic traffic agents, autonomous UAV navigation, and a multitask YOLOv8n perception module for detecting road defects, pedestrians, and vehicles while classifying road-defect subtypes. After synthetic-domain fine-tuning, the perception model achieved 0.959 mAP@0.5 and 0.940 macro F1-score on a held-out synthetic test set generated from the simulator. The digital twin was then used to evaluate hover-and-recheck, micro-repositioning, and skip-and-revisit recovery strategies across different traffic densities and flight altitudes. Results show that flight altitude strongly affects inspection coverage, while recovery strategies introduce different trade-offs between coverage, mission duration, energy consumption, and revisit behaviour. These findings demonstrate that digital twins can support the development and evaluation of traffic-aware UAV inspection strategies before real-world deployment. The full implementation and trained models are available at https://github.com/EdwinTSalcedo/RDMO-DigitalTwin. --- ### VeriBound: PAC-Bayesian Generalization Bounds for Process Reward Models Trained with Formal Verification Tools Authors: Amirul Rahman, Mohammed Sabih Alsharari Published: 2026-06-17 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2606.20740v1 Process Reward Models (PRMs) provide step-level verification for Large Language Model (LLM) reasoning, yet their training data acquisition remains a bottleneck: human annotation is costly and Monte Carlo roll-out estimates are noisy. A recent approach, FOVER, trains PRMs on step-level error labels automatically annotated by formal verification tools such as Z3 and Isabelle, and empirically observes cross-task generalization from symbolic tasks to diverse reasoning benchmarks. However, this generalization phenomenon lacks any theoretical explanation, and no formal bounds exist on the generalization error, sample complexity, convergence rate, or downstream Best-of-K performance of such PRMs. We propose VeriBound, a theoretical framework that provides PAC-Bayesian generalization bounds for PRMs trained with formal verification tools. We establish four main results: (i) a PAC-Bayesian generalization bound that relates the empirical verification error on formal-verification-annotated training data to the expected error on unseen reasoning tasks, with the bound depending on the formal verification accuracy and the divergence between training and test task distributions; (ii) a sample complexity result showing that $O(d \log(d/δ) / ε^2)$ formal-verification-annotated examples suffice to achieve generalization error $ε$ with probability $1-δ$, where $d$ is the complexity of the PRM hypothesis class; (iii) a convergence analysis proving that PRM training with formal verification labels converges at a linear rate under $L$-smoothness and bounded variance conditions; and (iv) an error propagation bound that relates step-level verification error to Best-of-K performance degradation. --- ### An approach with Visual and Tabular Mamba to multimodal medical data using Mixed Fusion Authors: Matheus B. Rocha, Gustavo B. Dettogni, Renato A. Krohling Published: 2026-06-17 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.20738v1 This article presents a complementary approach for integrating multimodal medical data in cancer classification, based on state space models represented by the Mamba architecture. To this end, a mixed multimodal fusion architecture, called Mixed Fusion, was employed and developed to enhance the interpretability of the decision-making process. The proposed approach explores two variants of Mamba: one dedicated to visual processing, responsible for classifying the lesion image and generating probabilities associated with the target classes, and another focused on tabular processing, which uses these probabilities together with clinical and/or sociodemographic data to produce the final diagnosis. The experiments were conducted on two medical datasets: PAD-UFES-20, composed of clinical images and information associated with skin lesions, and NDB-UFES, consisting of histopathological images and sociodemographic data related to oral cancer. The results indicate slightly lower performance in balanced accuracy, compared with Transformer-based approaches, on PAD-UFES-20, and superior performance on NDB-UFES. Additionally, substantial gains were observed in the recall metric. Furthermore, the adoption of the Mixed Fusion architecture enables the application of the Shapley Additive Explanations (SHAP) method, increasing the interpretability of the results. These findings indicate that Mamba-based models constitute a suitable alternative for multimodal classification in medical data, especially in scenarios in which sensitivity is a relevant requirement. --- ### Repeated Shared Access Enables Grokking, but Edit Propagation Depends on an Addressable Memory Authors: Yanan Niu Published: 2026-06-17 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.20737v2 We study factual edit propagation in a controlled synthetic knowledge-graph QA setting using a 2x2 grid that crosses loop recurrence with shared-memory access: a dense transformer (Dense), a looped transformer (Loop), a dense backbone with shared memory (Dense+Mem), and a looped backbone with shared memory (loop-memory coupling, LMC). The two factors dissociate. For learning, both routes to repeated shared access -- looped recomputation and repeated memory rereading -- cross the out-of-distribution (OOD) grokking barrier that Dense fails, so repeated shared access is the behavioral regularity, not a specific architecture. For editing, the substrates split along a different axis: applying a single localized factual edit (conditioned on direct success) and measuring 2-hop propagation on a shared pre-edit-correct set, the edit propagates strongly in both memory-bearing cells (LMC 0.78-0.92, Dense+Mem 0.71-0.96) and only weakly in the memory-free ones (Loop 0.04-0.30, Dense 0.00-0.03). The split is along the memory axis, not the loop axis: every memory-bearing seed exceeds every memory-free seed, with no detectable difference between the two memory cells. Crucially Dense+Mem has no recurrence, so the propagating ingredient is an addressable site that an edit can write to and later computation rereads, not loop recomputation; Loop is at best a partial intermediate. The affordance survives coarsening the store (N=128 to N=13): propagation attenuates but the memory/no-memory split persists, so fine granularity buys precision rather than the affordance itself. These results dissociate learning competence from editing affordance -- repeated shared access suffices to grok, but edit propagation depends on whether the substrate exposes an addressable memory that the forward computation can write to and later reread, an affordance that loop recurrence provides only partially. --- ### REKEY: Metadata-Grounded Visual-Key Regeneration for Contamination-Resilient VQA Evaluation Authors: Tengjie Lin, Yutao Sun, Jingwei Ni, Shuhan Ge, Hao-Xuan Ma, Yanting Miao, Wangyue Lu, Mingshuai Chen, Tiancheng Zhao, Jianwei Yin Published: 2026-06-17 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.20736v1 Static visual question answering (VQA) benchmarks age quickly: Once the items leak into training corpora, scores can reflect memorization rather than genuine visual ability, thus obscuring real progress. Rebuilding high-quality benchmarks such as V*Bench requires substantial human annotation, yet each static release can quickly become another leaked artifact. We propose ReKey, a live benchmark protocol that randomly regenerates the answer-bearing local detail, or visual key, in real images at evaluation time. Using human-validated edit slots, ReKey samples fresh instances with new answers, construction-grounded labels, and controlled visual-search difficulty. On V*Bench, the ReKey regenerated benchmark reveals a sharp score jump across eight frontier vision-language models (VLMs): The original items score 9.5--18.8 percentage points higher than the regenerated variants. By making the visual key renewable, ReKey keeps evaluation fresh as models and training data evolve. --- ### Robust Zero-Shot Generalization for Open-Vocabulary Action Recognition via Task Arithmetic Authors: Francesca Morandi, Omayma Moussadek, Federico Venturini, Mauro Suardi, Alessandro Banzatti, Francesco Cannarile, Angelo Porrello, Simone Calderara Published: 2026-06-17 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.20734v1 Open Vocabulary Action Recognition (OVAR) enables the recognition of novel actions by leveraging vision-language representations, overcoming the limitations of traditional closed-set approaches. However, achieving robust performance in real-world scenarios typically requires domain-specific fine-tuning, which is often costly and raises privacy and regulatory concerns. In this work, we propose an alternative paradigm that bypasses target-domain training and recombines knowledge from existing datasets and models. Leveraging model merging and task arithmetic, we extract and combine task vectors from models fine-tuned on diverse public OVAR datasets. We show that, in out-of-distribution settings, the resulting merged model achieves superior zero-shot generalization to the pre-trained base model. Code is available at https://github.com/omaymaMoussadek/robust-ovar --- ### XmoPipe: A Pipeline for Large-Scale In-the-Wild Human Motion Dataset Construction Authors: Nathan Salazar, Emmanuel Dellandréa, Mathieu Lefort, Alexandre Meyer Published: 2026-06-17 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.20731v1 Large-scale human motion datasets are essential for training robust motion models for analysis, synthesis, and understanding. While marker-based motion capture provides precise data, it is costly and limited in scale and diversity. Recent advances in monocular motion capture and video-language understanding open the way to extract plausible motion from unconstrained online videos. We present a scalable pipeline for constructing in-the-wild human motion datasets. From a few keywords, the system retrieves videos, extracts 3D body and facial motion, and generates high-level textual descriptions. The pipeline is flexible, enabling targeted collection of various motions, multi-person interactions, or expressive behaviors. We demonstrate its quality by training motion reconstruction and motion generation models, showing performance comparable to models trained on traditional motion capture datasets and strong cross-dataset generalization. --- ### LLM-Guided Test-Time Discovery of Quantum-Chemical Approximation Algorithms Authors: Masaya Hagai, Yuta Suzuki, Tomoya Murata, Shuhei Kurita, Masaki Adachi Published: 2026-06-17 Categories: physics.chem-ph, cond-mat.mtrl-sci, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.20729v1 Quantum chemistry simulations underpin modern materials discovery, yet their impact is limited by steep computational cost and dependence on fixed approximation schemes. Foundation models, such as machine-learned interatomic potentials, have accelerated parts of this workflow, but their reliance on large-scale pretraining restricts adaptability at the frontier of chemical space, where methodological innovation and sparse data are the norm. Agentic AI systems can automate existing simulation pipelines, yet they remain constrained by the predefined tools and algorithms they orchestrate. In response, we introduce LADeQ, an LLM-guided workflow that discovers, implements, and benchmarks candidate approximation algorithms at test-time within existing quantum chemistry codes. Rather than selecting from a predefined repertoire, LADeQ constructs candidate approximation schemes on demand, drawing on techniques from disciplines such as spatial statistics, circuit simulation, and kernel methods that have had little prior presence in electronic-structure theory. Because it builds on an out-of-the-box language model, LADeQ requires no task-specific pretraining or curated data, and the resulting implementations are transparent and inspectable, with explicitly traceable approximation errors that enable principled control of accuracy--efficiency trade-offs. We show that LADeQ accelerates coupled cluster singles and doubles (CCSD) and configuration interaction singles and doubles (CISD) calculations while keeping correlation-energy errors within user-specified tolerances, demonstrating autonomous, objective-driven discovery of approximation algorithms inside existing electronic-structure solvers. --- ### VTOS: Learning to Orchestrate Vision Tools by Co-Searching Solutions and Observers Authors: Jinchao Ge, Lingqiao Liu, Shuwen Zhao, Lei Wang Published: 2026-06-17 Categories: cs.CV, cs.CL arXiv: https://arxiv.org/abs/2606.20728v1 Vision foundation tools such as open-vocabulary detectors, segmentation models, and post-processing operators are powerful building blocks for computer vision, but their effectiveness depends heavily on how they are orchestrated: which tools are used, in what order, with what parameters, and under what visual conditions. Existing visual-programming agents typically generate a fixed solution pipeline, making them brittle under dense objects, occlusion, small targets, and domain shift. We introduce VTOS (Vision Tools Orchestration Search), a framework for adaptive visual tool orchestration through joint solution--observer search. VTOS co-searches executable solution programs that compose vision tools such as Grounding DINO, SAM, NMS, and slice-and-detect, together with observer programs that diagnose candidate solutions, identify failure modes, and generate actionable feedback. These observations are accumulated in a shared VisionThoughts knowledge base to guide subsequent search. We evaluate VTOS through two case studies: dense object counting on LVIS-Count and zero-shot plant-disease segmentation on PlantSeg-OOD, which stress different orchestration challenges including threshold calibration, NMS, slicing, mask refinement, and domain generalization. Across both tasks, VTOS outperforms static tool pipelines and agentic visual-programming baselines, showing that co-searching solutions and observers is an effective strategy for adapting vision tools to challenging computer vision tasks. --- ### FPGA-Accelerated Neuromorphic Vision System for Real-Time Orbital Object Detection Authors: Diego Hernández, Sebastián Valdivia, Vicente Westerhout, Esteban Vera, Daniel Yunge Published: 2026-06-17 Categories: cs.AR, cs.LG, cs.NE arXiv: https://arxiv.org/abs/2606.20727v1 The escalating congestion in orbital space demands advanced monitoring solutions. This work presents a comprehensive open-source framework for neuromorphic resident space object (RSO) detection, adapting the foundational grid clustering algorithm for FPGA acceleration. The system integrates a single event-based camera (EBC) with a custom, distributed processing architecture, where rapid spatial quantization is executed in programmable logic (FPGA) and cluster formation is managed by a software client. We validate this architecture through systematic sampling of night-sky observations from the EVAS dataset, demonstrating 97% detection accuracy for RSOs. The implementation, which serves as a foundational toolkit for event-based FPGA processing, achieves efficient throughput with a total power consumption of 8.5 W and deterministic processing latencies below 62 ms. The architecture's energy efficiency and high-precision detection position it as a viable solution for distributed space surveillance networks. --- ### How Well Can Your Video Model Remember? Measuring Memory-Budget Trade-offs in Long Video Understanding Authors: Yixian Tian Published: 2026-06-17 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.20726v1 We introduce a compact empirical model that quantifies how answer accuracy degrades as a function of frame budget B and temporal distance D in long video understanding -- analyzing performance when recalling content from D seconds in the past using a fraction B of total frames. Long-form models operate under strict budgets, yet no prior framework predicts how accuracy degrades as B shrinks and events recede. We fit a weighted least-squares model on ~155,000 binary predictions across ten models and three sampling strategies, deriving a law where logit-accuracy scales linearly in log-budget with a distance-dependent exponent that decays log-linearly with distance. This budget exponent α(D) captures the marginal value of extra frames at distance D. The law achieves cell-level weighted R^2 = 0.05-0.75 across models. Notably, budget effectiveness at D = 1000 s differs by \approx 7.4\times between the best streaming and base models. STREAMINGVLM achieves α(1000) = 1.26 (95% CI: (1.06, 1.58)), meaning a tenfold budget increase substantially improves long-distance accuracy, while the best Qwen3-VL base model reaches only α(1000) = 0.17 (CI: (0.04, 0.34)). In accuracy space, a 10\times budget increase at D = 1000 s yields +29 percentage points for STREAMINGVLM versus +4 pp for the base model. Sampling strategies show model-dependent trade-offs: random sampling yields higher base sensitivity but steeper distance decay. We demonstrate how α(D) enables principled budget allocation, including a model-ranking reversal at long distance, and propose it as a diagnostic metric for streaming video models. --- ### D2HDMap: Non-visible Driveline Map Prior for Online Vectorized HD Map Prediction Authors: Seojun Shon, Chikao Tsuchiya, Dhaval Bhanderi, David Ilstrup, Hsinmin Cheng, Christopher Ostafew Published: 2026-06-17 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.20725v1 Accurate, up-to-date representations of road structures are critical for the safe operation of autonomous vehicles. Existing systems rely either on costly, maintenance-heavy high-definition (HD) maps which compromise safety when outdated, or purely sensor-based online mapping which struggles with long-range reliability and occlusion. Systems incorporating map prior information into online mapping seek to overcome drawbacks of both approaches by combining them in some way. We propose 'Driveline To HD Map' (D2HDMap), an online mapping system that injects a lightweight, non-visible driveline prior to guide the estimation of visible road structures such as lane dividers, road boundaries and crosswalks. This prior incurs less effort to create and update compared to full HD map priors used in other approaches. We also show that training with such a prior can improve generalization at inference time when no prior is available. Ablation studies conducted on the nuScenes and Argoverse 2 dataset demonstrate that models trained using a driveline prior largely retain performance even when priors are not available. On a geographically disjoint split, D2HDMap achieves 44.8 mAP, surpassing recent state-of-the-art. Additionally, noise-aware training substantially increases robustness to realistic localization error. --- ### Cinematic Compositing Using Character-Environment-Harmonized Video Generation Models Authors: Tianyi Xiang, Mingming He, Li Ma, Jing Liao Published: 2026-06-17 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.20233v2 Cinematic compositing aims to integrate green-screen characters into novel environments while maintaining physical and photometric realism. Previous methods often fail to capture the complex bidirectional interactions between characters and their surroundings, which we characterize as Character-to-Environment (C2E) physical interaction and Environment-to-Character (E2C) lighting harmonization. To address this, we propose an end-to-end video diffusion framework that jointly models C2E and E2C interactions, specifically handling the challenges of interactive props. Our approach introduces a tri-mask-guided architecture with RGB-D joint denoising to ensure physically consistent interactions among the character, props, and environment. We further develop an efficient prior-driven data curation pipeline to construct high-quality relighting pairs without expensive rendering. Finally, a reference-conditioned mechanism enables controllable environment synthesis and precise prop replacement. Extensive experiments demonstrate that our framework significantly outperforms existing methods in cinematic-quality dynamic video compositing. --- ### A Layered Security Framework Against Prompt Injection in RAG-Based Chatbots Authors: Gulshan Saleem, Nisar Ahmed, Muhammad Imran Zaman, Ali Hassan Published: 2026-06-17 Categories: cs.CR, cs.CL arXiv: https://arxiv.org/abs/2606.19660v1 Prompt injection is ranked as the most critical vulnerability in large language model (LLM) deployments by the OWASP Top 10 for LLM Applications, yet existing defenses operate at isolated pipeline stages and remain incomplete. Input filters cannot inspect retrieved documents, while output monitors cannot prevent malicious payloads from reaching the model. Consequently, retrieval-augmented generation (RAG) chatbots remain vulnerable to indirect injection, where a poisoned knowledge-base document compromises every user whose query retrieves it. We present a three-layer framework that intercepts both direct and indirect prompt injection throughout the inference pipeline. Layer 1 screens user input using a rule-based pattern library and a fine-tuned semantic anomaly classifier. Layer 2 enforces a provenance-based instruction hierarchy during context assembly, preventing retrieved content from overriding operator policy. Layer 3 audits model output using a policy rule engine and semantic drift detector before delivery. A continuous audit loop aggregates structured logs and supports retraining to adapt the classifier to emerging attack patterns. The framework is model-agnostic and deploys as middleware without modifying the underlying LLM. Evaluation on 5,080 samples across GPT-4o, Llama 3, and Mistral 7B shows that the framework reduces Attack Success Rate (ASR) from 71.4\% to 11.3\%, outperforming the best single-layer baseline by 27.3 percentage points and a published guardrail system by 23.8 percentage points, while maintaining a 4.8\% false positive rate and a median latency overhead of 61.2 ms. Ablation studies confirm that all three layers provide complementary protection and that their combined effect exceeds the sum of individual contributions. --- ### SAGE-OPD: Selective Agent-Guided Intervention for Multi-Turn On-Policy Distillation Authors: Yuhang Zhou, Lizhu Zhang, Yifan Wu, Mingyi Wang, Bo Peng, Jiayi Liu, Xiangjun Fan, Zhuokai Zhao Published: 2026-06-17 Categories: cs.CL arXiv: https://arxiv.org/abs/2606.19659v1 On-policy distillation (OPD) improves student models by training them on trajectories induced by their own policy, making it a promising approach for mitigating exposure bias in agent training. However, most OPD studies focus on single-turn settings, while realistic LLM agents interact with environments over multiple turns. In this regime, early errors can alter future observations and compound across the trajectory, and standard dense token-level OPD becomes brittle, as it may over-penalize semantically valid alternatives, reinforce local degeneracies such as repeated actions, and propagate unreliable teacher supervision on off-distribution histories. We propose SAGE-OPD, a verifier-free selective intervention framework specifically designed for multi-turn OPD. Instead of applying teacher supervision uniformly across all turns, SAGE-OPD first observes environment feedback and uses teacher judgment to decide whether each student response should be skipped or intervened on. To further address compounding errors, SAGE-OPD weights token-level distillation by teacher confidence, reducing the influence of uncertain teacher distributions on corrupted or ambiguous histories. Finally, SAGE-OPD applies loss normalization to preserve the overall loss scale of standard OPD while retaining selective turn-level weighting. Experiments on agent tasks show that SAGE-OPD consistently improves over baselines, achieving up to a 13.3% relative improvement in ALFWorld unseen success rate over standard OPD. Ablation studies further demonstrate that turn-level intervention, teacher confidence weighting, and loss normalization provide complementary benefits. Our results suggest that effective multi-turn OPD should remain on-policy, but teacher supervision should be selectively allocated to turns where intervention is necessary and reliable. --- ### Denoising Implicit Feedback for Cold-start Recommendation Authors: Gaode Chen, Shicheng Wang, Shikun Li, Rui Huang, Xinghua Zhang, Yunze Luo, Shipeng Li, Shiming Ge, Ruina Sun, Yinjie Jiang, Jun Zhang Published: 2026-06-17 Categories: cs.AI, cs.IR, cs.MM arXiv: https://arxiv.org/abs/2606.19658v1 Implicit feedback is widely used in recommender systems due to its accessibility and generality, yet it usually presents noisy samples (e.g., clickbait, position bias). Meanwhile, recommenders inevitably face the item cold-start problem due to the continuous influx of new items. We identify that cold items are more prone to noisy samples due to the aforementioned factors, and researchers often overlook the significance of denoising implicit feedback for cold items. Previous denoising studies usually identify noisy samples based on heuristic patterns, such as higher loss values, and mitigate noise through sample selection or re-weighting. However, these methods have limited adaptability and are ineffective in cold-start scenarios. To achieve denoising implicit feedback for cold-start recommendation, we propose a model-agnostic denoising method called DIF. First, user preferences for content remain stable, which allows us to infer pseudo-labels indicating whether a user is interested in a cold item through content-similar warm items. Furthermore, to improve pseudo-label accuracy, we model the confidence of pseudo-labels based on the content similarity between the cold item and warm items, and then aggregate multiple pseudo-labels for each sample. Finally, we explicitly estimate the uncertainty of the noisy sample label by considering its relative entropy and the cold-start status of the item, which adaptively guides the role of pseudo-labels to correct the noisy labels at the sample level. DIF's superiority is supported by both theoretical justification and extensive experiments on real-world datasets. The method has been deployed on a billion-user scale short video application Kuaishou and has significantly improved various commercial metrics within cold-start scenarios. --- ### DF-ExpEnse: Diffusion Filtered Exploration for Sample Efficient Finetuning Authors: Calvin Luo, Chen Sun, Shuran Song Published: 2026-06-17 Categories: cs.RO, cs.LG arXiv: https://arxiv.org/abs/2606.19656v1 A natural recipe for intelligent robotic decision-making is initializing from pretrained generative control policies, which have summarized offline experience, and adapting them to self-collected online experience. We present DF-ExpEnse, an exploration technique that improves the quality of online experience collection, thus increasing finetuning sample-efficiency. DF-ExpEnse leverages the multimodal modeling capabilities of the generative control policy to create an expressive and tractably evaluatable candidate set. It then utilizes an ensemble of critics to identify the action that best balances quality with high exploration interest. In fleet settings, DF-ExpEnse further enables cross-agent communication to facilitate collaborative exploration as a group. DF-ExpEnse can be seamlessly integrated with existing strategies that finetune pretrained generative control policies via reinforcement learning. We experimentally validate consistent sample-efficiency benefits through DF-ExpEnse across a variety of manipulation and locomotion tasks, compared to default finetuning and alternative action selection schemes. Project can be found at https://df-expense.github.io. --- ### Convex training of Lipschitz-regularized shallow neural networks Authors: Chao Yin, Antoine Lesage-Landry Published: 2026-06-17 Categories: cs.LG arXiv: https://arxiv.org/abs/2606.19652v1 In this work, we introduce a training procedure for shallow neural networks that promotes robustness against adversarial attacks. We solve a non-convex Lipschitz-regularized training program by introducing a convex restriction that can be efficiently solved to global optimality. Our approach can be employed as a post-processing step by taking a pre-trained network as an initial solution to then solving the convex program whose optimal network is guaranteed to be no worse than the initial one. We illustrate the improvements of our training procedure with experiments using real world datasets for regression tasks under an adversarial setting. We show numerically that solving our proposed convex program yields networks with lower objective values on the Lipschitz-regularized program compared to existing methods. Additionally, we show that on certain datasets, networks obtained using our convex training program are both more accurate and robust with respect to adversarial attacks. --- ### BrainG3N: A Dual-Purpose Tokenizer for Controllable 3D Brain MRI Generation Authors: Max Van Puyvelde, Ibrahim Gulluk, Wim Van Criekinge, Olivier Gevaert Published: 2026-06-17 Categories: cs.AI, cs.CV, cs.LG arXiv: https://arxiv.org/abs/2606.19651v2 Three-dimensional (3D) brain MRI is central to clinical neurology and neuro-oncology, where generative models could augment under-represented cohorts, simulate disease trajectories, and support privacy-preserving data sharing. Latent diffusion has been the go-to solution for modeling imaging data, but it places two competing demands on the tokenizer: encoder embeddings must retain the clinical information that downstream tasks act on, and the decoder must reconstruct anatomically faithful volumes. Existing reconstruction-driven tokenizers achieve the second at the expense of the first. To address this, we introduce a fully volumetric masked-autoencoder (MAE) based tokenizer for 3D brain MRI latent diffusion, decoupling encoder and decoder: a frozen 3D MAE encoder produces clinically informative embeddings, while a dedicated CNN decoder reconstructs voxels from a linear projection of those embeddings. We pretrain the encoder on 35,309 volumes from 18 public cohorts spanning four modalities, ten disease categories, and 200+ acquisition sites, and demonstrate its dual utility in two settings. First, on a 23-task linear-probing benchmark, the encoder outperforms or matches SOTA models (i.e., BrainIAC, BrainSegFounder, and MedicalNet) on 21 of 23 tasks. Second, a conditional diffusion transformer (DiT) trained on these clinically informative embeddings supports both conditional generation across six variables and patient-specific longitudinal forecasting. Together these results establish a single 3D brain-MRI embedding space capable of both downstream clinical tasks and controllable generation. --- ### From 50K to 8.2 Million in 24 Hours: Vozinha's Algorithmic Consecration and the Multilingual Making of World Cup Visibility Authors: Vinicius Covas Published: 2026-06-17 Categories: cs.CL, cs.CY, cs.SI arXiv: https://arxiv.org/abs/2606.19647v1 We present a multilingual computational discourse analysis of how language constructed the algorithmic consecration of Vozinha, the 40-year-old Cape Verde goalkeeper, after Spain 0-0 Cape Verde at the 2026 FIFA World Cup. The study contributes a multilingual corpus in Portuguese, Spanish, English, and French; a nine-frame narrative taxonomy with cue-based frame annotation; a reproducible annotation pipeline combining LLM-assisted suggestion with human validation; and an analysis of cross-lingual narrative diffusion across discourse phases. We treat the platform follower count itself, narrated as "50k to 8M", as a linguistic object: a circulating and narratable proof of visibility rather than a mere measurement. The follower-growth timeline is used only as contextual metadata: we reconstruct a conservative phase structure, not a continuous API-native series, and type every datapoint by value class, confidence, and evidence type. The only exact primary scraper anchor is 8,235,652 followers at 2026-06-16 15:47 UTC; all other figures are reported as estimated ranges or thresholds, including an estimated pre-match baseline of 45k-56k. Findings suggest that distinct languages carried distinct frames: Portuguese mobilization, Spanish crisis, English nation-making, and a shared platform-metric spectacle through which peripheral athletic performance became globally visible. As a v0.1 pilot, the paper releases the corpus schema, frame taxonomy, annotation guidelines, hashed visual-evidence log, and typed timeline, while flagging full double annotation and inter-annotator agreement as planned work. --- ### SAFE-Cascade: Cost-Adaptive Vision-Language Routing for Chart Question Answering Authors: Ayush Dwivedi, Qixin Wang, Ashvi Soni, Ruoteng Wang, Han Li, Animesh Mahapatra, Neeraj Agrawal, Xintao Wu Published: 2026-06-17 Categories: cs.IR, cs.CV arXiv: https://arxiv.org/abs/2606.19646v1 Vision-language models (VLMs) are powerful for chart question answering, but invoking a VLM for every query can be unnecessarily expensive when many questions are answerable from OCR text and lightweight language reasoning. We demonstrate SAFE-Cascade, an interactive system for cost-adaptive chart question answering. Given a chart image and a natural-language question, SAFE-Cascade first extracts chart text with OCR, obtains a provisional answer from a text-only language model, and then uses a learned router to decide whether to accept the text answer or escalate to a VLM. The demo exposes this decision process to users: OCR evidence, text-only answer, routing probability, escalation decision, final answer, estimated cost, and estimated latency are shown side by side. SAFE-Cascade is designed as a transparent interface for understanding when visual grounding is actually needed. Users can upload or select charts, ask questions, inspect the evidence used by each pathway, compare text-only and VLM answers, and adjust the escalation threshold to explore the accuracy-cost frontier. The system is implemented with Azure Document Intelligence for OCR, gpt-5-mini as the text-only model, gemini-2.5-flash-image as the VLM, and a Random Forest router trained on inference-time features. On a held-out ChartQA test split of 375 examples from a 2,500-example experiment, SAFE-Cascade achieves 69.1% unified accuracy with 73.1% VLM invocation, compared with 67.7% accuracy and 100% VLM invocation for the full-VLM baseline. The observed +1.4 percentage-point difference is statistically uncertain, so we interpret SAFE-Cascade as matching full-VLM performance while reducing VLM calls by 26.9% and estimated cost by 9.3%. The demonstration shows how selective modality routing can make multimodal knowledge systems more transparent, tunable, and cost-aware. --- ### Variational Consensus Monte Carlo for Bayesian Mixture Authors: Julie Fendler, Francesca L. Crowe, Tom Marshall, Sylvia Richardson, Paul D. W. Kirk Published: 2026-06-17 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2606.19643v1 Motivated by the privacy, sensitivity and sharing limitations of health data, we present a comprehensive pipeline for inference of Bayesian mixture models within a federated learning setting, i.e. when data cannot be fully shared or pooled across compute nodes. We adopt a Consensus Monte Carlo (CMC) approach, in which an MCMC algorithm is run independently within each data silo to estimate local posterior distributions, which are then aggregated to approximate the posterior over the full data. The variational CMC approach of Rabinovich, Angelino and Jordan (2015) (1) frames the aggregation step as a variational inference problem, but their application to mixtures assumes the number of clusters and key mixture parameters to be known. Our main methodological contributions are: (i) an extension of variational CMC to over-fitted Bayesian mixture models that infer the number of clusters and all model parameters, without requiring conjugacy; (ii) novel cluster-matching algorithms suitable for cross-silo settings in which not every cluster appears in each local dataset; (iii) a number of inference strategies for the aggregation step, matched to different federated learning constraints; and (iv) guidelines for choosing among these in practice. A comprehensive simulation study validates the framework and allows us to compare to state-of-the-art federated learning alternatives. Notably, we show that when the composition of local datasets reflects the underlying clustering structure in the data, our approach can recover small clusters with greater accuracy than standard MCMC applied to the pooled data. We illustrate the framework on large-scale electronic health record data, identifying multi-morbidity patterns in a British geriatric population. --- ### Rigorous uncertainty quantification of probabilistic AI weather forecasts with conformal prediction Authors: Anna Asch, Raphael Rossellini, Pedram Hassanzadeh, Rebecca Willett Published: 2026-06-17 Categories: physics.ao-ph, stat.AP, stat.ML arXiv: https://arxiv.org/abs/2606.19642v1 Probabilistic weather forecasting is undergoing rapid transformation with artificial intelligence (AI). In traditional numerical weather prediction, computing power can limit how well ensemble forecasts approximate the unknown statistical distribution of future states. AI models facilitate larger ensembles and are trained with probabilistic considerations, ideally leading to better uncertainty quantification. Forecasts from these state-of-the-art models are often considered well-calibrated. However, here we show that the statistical coverage of such models, the ultimate measure of calibration, can struggle, especially on extreme events. To address this shortcoming, we employ conformal prediction, a class of statistical methods that mathematically guarantees coverage under no distributional assumptions, unlike previous post-processing techniques. We apply online conformal prediction to temperature and precipitation forecasts (including extremes) of three leading global weather models, GenCast, NeuralGCM, and AIFS-ENS, ensuring calibrated uncertainty at no expense to other probabilistic metrics. This post-processing method can be applied to any forecasting model. --- ### Scaling Self-Play for End-to-End Driving Authors: Luke Rowe, Roger Girgis, Rodrigue de Schaetzen, Daphne Cornelisse, Alaap Grandhi, Felix Heide, Eugene Vinitsky, Christopher Pal, Liam Paull Published: 2026-06-17 Categories: cs.RO, cs.CV arXiv: https://arxiv.org/abs/2606.19641v2 End-to-end autonomous driving models are typically trained on offline human-demonstration datasets that provide limited state coverage and often no closed-loop feedback, making them prone to compounding errors when deployed in closed-loop and brittle to long-tail agent interactions. To overcome these limitations, we propose an alternative strategy for training end-to-end driving models: large-scale self-play directly from pixels in simulation. While prior self-play approaches have shown promising transfer to real-world driving, they typically assume vectorized Bird's-Eye-View (BEV) observations that are incompatible with end-to-end policies operating directly on sensor observations. To this end, we introduce Gigapixel, a high-throughput batched driving simulator with perspective rendering, enabling scalable self-play directly from pixel observations. Rather than targeting compute-costly photorealistic sensor simulation, Gigapixel renders a simplified bounding-box world that preserves essential scene structure while achieving throughput at 50k agent steps per second. Since direct pixel-space self-play RL is prohibitively sample-inefficient at end-to-end model scale, we propose self-play DAgger training: we train pixel-based policies in self-play via on-policy distillation from a privileged RL teacher. To bridge the sim-to-real gap, we subsequently transfer the self-play trained policies to real-world sensor data through lightweight perception adaptation. Policies trained in Gigapixel and adapted to real-world sensor data achieve competitive performance on the HUGSIM and NAVSIM-v2 benchmarks without human trajectory supervision. Moreover, scaling self-play training yields proportional gains in policy performance, establishing self-play as a practical and scalable strategy for training end-to-end models. --- ### Creating Multilingual Mental Health Dialogue Datasets: Limits of Persona-Based Localization via Nationality and Language Authors: Yunkai Xu, Saeed Abdullah Published: 2026-06-17 Categories: cs.CL, cs.AI, cs.HC arXiv: https://arxiv.org/abs/2606.19640v1 AI and large language models (LLMs) have emerged as promising tools to address global mental health challenges. Despite the global nature of these challenges, there remains a critical shortage of high-quality datasets for training and evaluating such systems. To mitigate this gap, researchers increasingly generate synthetic clinical personas to simulate user data and test digital mental health support systems. However, most validated personas rely on English-centric contexts. This paper investigates whether similar persona-based methods can be used to generate multilingual mental health datasets. We modified nationality and language parameters in personas to generate clinical dialogues in Mandarin, Bengali, and Hindi. We then examined how different LLMs perform when evaluating the depression severity of these generated multilingual datasets against the baseline in English. Our findings indicate that just adding nationality and language parameters in personas might not be adequate, as it can introduce clinical inconsistency across languages. LLM judge models often exhibit inaccuracies in assessing depression severity in non-English texts, with performance varying across different models. This exposes the systemic limitations of applying English-centric personas to multilingual contexts. Ultimately, our work highlights the urgent need for culturally responsive data generation to ensure equitable mental health systems globally. --- ### MiqraBERT: Regression-Based Sentence-BERT Finetuning for Biblical Hebrew Parallel Detection Authors: David M. Smiley Published: 2026-06-17 Categories: cs.CL arXiv: https://arxiv.org/abs/2606.19638v1 Textual reuse pervades the Hebrew Bible, yet the computational methods used to detect it still rest largely on lexical overlap, and they falter once a parallel involves paraphrase, lexical substitution, or syntactic reworking. This paper introduces MiqraBERT, a Sentence-BERT model finetuned from AlephBERT (a Modern Hebrew encoder) for verse-level semantic similarity in Biblical Hebrew. The training set comprises 1,650 labeled verse and half-verse pairs: 825 true parallels drawn from the Chronicles synoptic material and from foundational studies of poetic parallelism, balanced against 825 randomly sampled negatives. Through cosine-similarity regression, the model learns an embedding space in which parallel verses cluster together and unrelated verses move apart. We evaluate separation with distribution-based metrics, Wasserstein distance and the overlap coefficient, across ten random seeds. MiqraBERT improves distributional separation 2.7-fold over the pre-trained baseline and reduces the ambiguous overlap region from roughly 24% to about 6%. Narrative synoptic parallels reach a recall@10 of 87.1%; poetic parallels remain difficult, below 9%. This genre-dependent asymmetry confines the model's reliable scope to narrative textual reuse. MiqraBERT is publicly available at https://huggingface.co/davidmsmiley/MiqraBERT --- ### Before the Labels: How Dataset Construction Shapes Suicidality Detection in Clinical Text Authors: Priyanshi Garg, Ishita Rao, Jieqiong Ding, Amandalynne Paullada Published: 2026-06-17 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2606.19637v1 Clinical NLP increasingly relies on electronic health record (EHR) data to detect suicidal behaviors, treating clinical documentation as more reliable ground truth than social media. We argue that this framing obscures how EHR-based suicidality datasets encode a particular operationalization of suicidality, shaped by who authors the data, how episodes are bounded, and how ambiguity is resolved. We ground this argument in a case study of the ScAN dataset, built over MIMIC-III clinical notes. We show how governance constraints, ICD-based cohort selection, single-annotator labeling, and hospital-stay-level aggregation produce labels that reflect clinician-documented judgments, treat suicidality as a bounded episode, and assume that intent can be reliably inferred from documentation. A linguistic analysis demonstrates that identical labels subsume heterogeneous clinical framings differing in temporality, negation, and uncertainty. We argue that clinical NLP should examine the assumptions embedded in suicidality datasets before interpreting their labels as ground truth. --- ### Hard or Just Unreached? Diagnosing the Sampling Blind Spot in Math-Reasoning Difficulty Estimation Authors: Luca Zhou, Sajel Shah, Emanuele Rodolà, Roberto Dessì Published: 2026-06-17 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2606.19636v3 Math and science reasoning benchmarks rely on pass@k, the fraction of sampled chains that reach gold, as the canonical per-example difficulty signal. The same signal drives RL with verifiable rewards, math data curation, synthetic curricula, and verifier training. We show this proxy has a persistent blind spot on its hardest stratum: on the eight free-form math cells we test (GSM8K and MATH across four open-weight models), 10.3-22.9% of the examples that no sampling seed solves in six tries are instead solved at matched compute by a six-chain deterministic regime. These are greedy decoding plus five cheap residual-stream perturbations applied via activation grafting, while greedy alone solves at most 6% on these math cells. Recovery scales with the additional budget, across perturbations whose mechanistic distinctness we verify across all twelve cells (cross-kind fix-set Jaccard <= 0.47 in every setup). Activation grafting is used as an intervention on internal representations, not a decoding method; we use it purely as a diagnostic and diversification tool, and our recovered items show that the pass@k= 0 % stratum is structurally identifiable in the residual stream rather than that the unmodified model reaches them under ordinary inference. --- ### Token Factory: Efficiently Integrating Diverse Signals into Large Recommendation Models Authors: Xilun Chen, Shao-Chuan Wang, Baykal Cakici, Lukasz Heldt, Lichan Hong, Raghu Keshavan, Aniruddh Nath, Li Wei, Xinyang Yi Published: 2026-06-17 Categories: cs.IR, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.19635v3 Large Recommendation Models (LRMs) have demonstrated promising capabilities in industry-scale recommendation tasks. However, holistically integrating traditional signals into these transformer-based architectures effectively and efficiently remains a major challenge. Conventional approaches that "textualize" these signals directly or create discrete item representations often lead to excessively long prompts, substantial memory footprints, and high computational overhead. To overcome these limitations, we propose "Token Factory", a framework designed to transform traditional signals into "soft tokens" that can be directly processed by LRMs. This approach enables efficient integration and compression of heterogeneous input features, preventing prompt length explosion while enhancing model performance. We detail the architecture of Token Factory and present experimental results validating its effectiveness in a production-scale recommendation environment. --- ### CTS-MoE: Implicit Terrain Adaptation via Mixture-of-Experts for Perceptive Locomotion Authors: Francisco Affonso, Matheus P. Angarola, Ana Luiza Mineiro, Aditya Potnis, Marcelo Becker, Girish Chowdhary Published: 2026-06-17 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2606.19633v1 Perceptive legged locomotion over discontinuous terrain (e.g., stairs, gaps, and obstacles) requires adaptive behavior, as a single conservative gait cannot produce the anticipatory maneuvers needed for abrupt topology changes. Cast as multi-task reinforcement learning, this problem introduces a tension between sharing and separation. Tasks use a common locomotion base but have conflicting rewards, so a policy must share behavior while avoiding value interference. Prior work addresses only one side, with monolithic policies sacrificing specialization and hierarchical sub-policies sacrificing generalization across transitions and unseen terrain. We propose CTS-MoE, which combines a dense mixture-of-experts actor with perception-based gating to compose shared behaviors and a multi-critic with task-specific value heads to prevent interference. The model is trained end-to-end in a single-stage concurrent teacher-student setup that handles partial observability and avoids sequential distillation, with task labels used only during training. At deployment, routing depends solely on perception, allowing terrain adaptation without a high-level selector or terrain classifier. Experiments on a Unitree Go1 in simulation and on hardware across seen and unseen terrains show task-aware specialization, with lower tracking error and higher success rates than monolithic baselines. Project Website: https://cts-moe.github.io/ . --- ### Formal Verification of Learned Multi-Agent Communication Policies via Decision Tree Distillation Authors: Ahmad Farooq, Kamran Iqbal Published: 2026-06-17 Categories: cs.RO, cs.AI, cs.LG, cs.LO, cs.MA arXiv: https://arxiv.org/abs/2606.19632v1 Multi-agent reinforcement learning (MARL) enables agents to develop coordination strategies through emergent communication, but neural policies lack the formal safety guarantees required for safety-critical robotic deployment in drone swarms and autonomous vehicle fleets. We present the first end-to-end framework for safety verification of learned multi-agent communication policies through policy abstraction: neural policies are distilled into interpretable decision trees, then formally verified, with empirical validation confirming that verified safety properties transfer to original networks. Our four-stage pipeline consists of domain-specific feature extraction from agent observations, decision tree distillation achieving 97.9% +/- 1.2% fidelity to neural policies, automated translation to PRISM probabilistic model checker specifications with complete feature-to-state-variable correspondence, and compositional verification of Probabilistic Computation Tree Logic (PCTL) properties via pairwise decomposition with union-bound aggregation and empirical neighbor modeling. Evaluating Vector-Quantized Variational Information Bottleneck (VQ-VIB) policies for multi-drone coordination with 5-7 agents, we verify 18 temporal logic properties across safety, liveness, and cooperation, achieving 88.9% property satisfaction with all five safety thresholds satisfied (0.3% collision probability vs. 1% threshold). Monte Carlo validation of original neural policies confirms that verified safety properties transfer with <=0.6 percentage-point deviation (95% CI). Discrete VQ-VIB messages provide +11.6 to +13.6 percentage-point fidelity advantages over continuous methods, enabling 3-4x faster verification. Our framework provides empirically validated safety verification for distilled policy abstractions, serving as a practical bridge between deep MARL and formal safety workflows for multi-robot deployment. --- ### AI4SE and SE4AI Exploration: A Decade Looking Back and Forward Authors: H. Sinan Bank, Daniel R. Herber, Thomas Bradley Published: 2026-06-17 Categories: cs.AI, cs.DL, eess.SY arXiv: https://arxiv.org/abs/2606.19630v1 The March 2020 INCOSE INSIGHT special issue on AI and Systems Engineering (SE) became the most downloaded issue in the publication's history and launched a research community that now draws over 250 registrants to its annual workshop. In this article, we trace the progress in AI and SE across three phases (labeled here foundational, applied, and LLM inflection) based on the authors' reading of the field's core papers, and describe our opinions of where the community has converged and where critical gaps remain. Separately, a human-AI agreement literature review leveraging both human expertise and six AI models was performed to assess the relevance of 1,712 INCOSE INSIGHT articles and 889 SERC publications. The results identify five critical research gaps and offer guidance for practitioners navigating AI adoption, assurance, and workforce transformation in SE. We share the agreement data and the AI4SE/SE4AI Explorer web application so readers can compare their own relevance judgments with the human and AI raters. --- ### RIVET: Robust Idempotent Voice Attribute Editing Authors: Dareen Alharthi, Bhuvan Koduru, Rita Singh, Bhiksha Raj Published: 2026-06-17 Categories: cs.SD, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.19629v1 Voice attribute editing models modify characteristics such as age and gender while preserving speaker identity. In large-scale speech datasets, however, attribute annotations are often noisy or inconsistent, which can cause conditional generative models to produce unstable edits. In this work, we show that idempotency provides an effective mechanism for improving robustness to noisy labels. An idempotent operator is one for which repeated application does not change the result, i.e., f(f(x)) = f(x). Enforcing this property acts as an implicit regularizer that reduces sensitivity to mislabeled examples. We introduce RIVET, a training framework that incorporates an idempotency objective to improve robustness to label noise. We evaluate RIVET under controlled label noise and on the GLOBE dataset with naturally noisy annotations. RIVET improves editing success and better preserves speaker identity than standard training, showing that idempotency improves robustness in voice editing models. --- ### VCG: A Multimodal Retrieval Framework for E-Commerce Video Feeds under Extreme Cold-Start Conditions Authors: Katya Mirylenka, Egor Malykh, Mahdyar Ravanbakhsh, Michael Gygli, Marco-Andrea Buchmann, Andrew Dzhoha, Svitlana Borzenko, Francesca Catino, Mohamed Gaafar, Maarten Versteegh, Thomas Kober, Dario d'Andrea et al. Published: 2026-06-17 Categories: cs.IR, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.19627v1 The digital commerce landscape is shifting from static, search-driven catalogs to dynamic, immersive video feeds. This transition introduces an ``extreme cold-start'' problem: unlike traditional items, new short-form videos lack the dense interaction history required for collaborative filtering. Furthermore, immersive feeds introduce strong position and duration biases that distort standard engagement signals. In this paper, we demonstrate the Video Candidate Generation (VCG) system, a scalable multimodal retrieval engine designed to solve these challenges in a large-scale e-commerce environment. By leveraging a domain-adapted vision-language model (based on CLIP), we map users and videos into a shared semantic space, enabling zero-shot retrieval based on visual content rather than behavioral history. We detail the system's architecture and present a rigorous evaluation comparing generative (LLM) vs. discriminative (CLIP) embeddings. Our results show that while generative models excel at attribute prediction, they suffer from embedding space collapse in retrieval tasks. Online A/B testing demonstrates that VCG effectively mitigates engagement biases, yielding a 50\% uplift in deep video completion. To showcase the system's capabilities, we present an interactive demonstration featuring three bi-directional retrieval scenarios: Product-to-Video, Video-to-Product, and Zero-Shot Semantic Search. --- ### Toten: A Knowledge-Based System For Structure-Preserving Representation Of Physical Quantities And Technical Notation In Brazilian Portuguese Authors: Antonio de Sousa Leitão Filho, Allan Kardec Duailibe Barros Filho, Fabrício Saul Lima. Selby Mykael Lima dos Santos, Rejani Bandeira Vieira Sousa Published: 2026-06-17 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2606.19626v2 AI pipelines that reason quantitatively over technical text depend on input where physical quantities, numbers, units, and symbolic expressions arrive intact; when these entities fragment at tokenization, errors propagate downstream. Byte-Pair Encoding, optimized for vocabulary compression, is blind to such entities and fragments them into arbitrary subwords -- a problem aggravated in technical Brazilian Portuguese. We present TOTEN, a knowledge-based system whose input representation preserves each technical entity as a whole, typed unit: vocabulary is not derived statistically but classified declaratively under a formal ontology of engineering entities (OEE). The core is the triple <O, classify, {inst_tau}>: types, principles, and invariants; a classifier mapping raw text into typed regions; and instantiators yielding a self-descriptive representation. Integrity rests on deterministic coupling to three external authorities: Pint (dimensional), Unicode Character Database (typographic), and RSLP (Portuguese morphology). We evaluate four properties verifiable by construction -- atomicity, dimensional equivalence, typographic robustness, numerical reconstruction -- on an internal benchmark (EngQuant, N=800) and four Brazilian Portuguese external corpora (N=1771 eligible cases), and report detection recall. Against eight state-of-the-art baselines, TOTEN reaches unit atomicity in all contrasts and reconstruction of 0.775-0.904 externally vs. 0.627-0.703 for the best (Quantulum3); on EngQuant, 0.780 vs. 0.340. Differences are significant (McNemar, Holm-corrected). Spearman correlation between internal and external rankings confirms concurrent validity of the control benchmark. TOTEN shows statistical parity with Pint in dimensional equivalence. The result is a structurally faithful, auditable, low-cost input layer for intelligent systems on technical knowledge, without generative models. --- ### Capability Provenance in Language Models: A Case Study in Social Reasoning Authors: Glenn Matlin, Chandreyi Chakraborty, Saehee Eom, Mika Okamoto, Rayan Castilla, Louis Jaburi, Alvin Deng, Taywon Min, Lucia Quirke, Stella Biderman, Mark Riedl Published: 2026-06-17 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2606.19625v4 We use training-data attribution as an interpretable tool for capability discovery, mapping which regions of the pretraining corpus support social reasoning versus STEM reasoning in OLMo3-7B. Training-data attribution measures how strongly each training document influences a model's predictions on a benchmark, but document-level scores are too noisy to identify which corpus regions support which capabilities. We compute gradient-based attribution (TrackStar via Bergson) over a working set drawn from the de-duplicated Dolma3 mix, aggregate influence across WebOrganizer's 24-format x 24-topic taxonomy (576 bins), and contrast benchmark pairs in a 2x2 design that varies domain (social vs. STEM) and capability type (reasoning vs. knowledge): SocialIQA and MMLU Social Sciences against ARC-Challenge and MMLU STEM. Social and STEM reasoning draw on qualitatively distinct corpus regions, and the contrast is sharper at the reasoning level than at the knowledge level. Targeted machine unlearning provides partial causal validation: forgetting high-attribution topics (e.g., Literature for SocialIQA) degrades the aligned benchmark more than within-topic random baselines. We open-source all code, data artifacts, influence scores, and checkpoints at https://github.com/HCAI-Lab-GT/capabilibara and https://huggingface.co/HCAI-Lab-GT. --- ### Capability Provenance in Language Models: A Case Study in Social Reasoning Authors: Glenn Matlin, Chandreyi Chakraborty, Saehee Eom, Mika Okamoto, Rayan Castilla, Louis Jaburi, Alvin Deng, Taywon Min, Lucia Quirke, Stella Biderman, Mark Riedl Published: 2026-06-17 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2606.19625v2 We use training-data attribution as an interpretable tool for capability discovery, mapping which regions of the pretraining corpus support social-reasoning versus STEM-reasoning in OLMo3-7B. Training-data attribution measures how strongly each training document influences a model's predictions on a benchmark, but document-level scores are too noisy to identify which corpus regions support which capabilities. We compute gradient-based attribution (TrackStar via Bergson) over a working set drawn from the de-duplicated Dolma3 mix, aggregate influence across WebOrganizer's 24-format x 24-topic taxonomy (576 bins), and contrast benchmark pairs in a 2x2 design that varies domain (social vs. STEM) and capability type (reasoning vs. knowledge): SocialIQA and MMLU Social Sciences against ARC-Challenge and MMLU STEM. Social and STEM reasoning draw on qualitatively distinct corpus regions, and the contrast is sharper at the reasoning level than at the knowledge level. Targeted machine unlearning provides partial causal validation: forgetting high-attribution topics (e.g., Literature for SocialIQA) degrades the aligned benchmark more than within-topic random baselines. We validate on other open-data model, Comma v0.1 7B-2T (Common Pile) and DCLM-Baseline-7B (DataComp-LM): causal selectivity holds on both models, while the provenance map is ecosystem-specific. We open-source all code, data artifacts, influence scores, and checkpoints at https://github.com/eilab-gt/capabilibara and https://huggingface.co/HCAI-Lab. --- ### MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery Authors: Hongxuan Liu, Roman Bushuiev, Ivy Lightheart, Mrunali Manjrekar, Anton Bushuiev, Magdalena Lederbauer, Filip Jozefov, Yinkai Wang, Soha Hassoun, Josef Sivic, James Taylor, Runzhong Wang et al. Published: 2026-06-17 Categories: cs.LG arXiv: https://arxiv.org/abs/2606.19624v1 Reliable benchmarking is critical for developing machine learning models for tandem mass spectrometry (MS/MS) based molecule discovery. Subtle issues in experimental design and model evaluation procedures can degrade the trustworthiness of such benchmarks and lead to erroneous conclusions. We conduct a thorough review of model evaluation issues in the recent MS/MS machine learning literature, using the standard MassSpecGym benchmark suite as a case study to illustrate the impact of these issues. We find evaluation issues in at least 17 of 26 papers reporting MassSpecGym benchmark results in the first year of its adoption. We isolate three classes of failures: (i) data leakage, (ii) shortcut learning, and (iii) implementation bugs and metric divergence. Through extensive experimentation and code replication, we quantify the impact of these issues and show how they corrupt the evaluation standards MassSpecGym was designed to enforce. We distill our findings into recommendations generalizable to MS/MS challenges, benchmarks, and custom evaluation setups. We also release MassSpecGym v1.5, an implementation of our recommendations in the MassSpecGym benchmarking suite which addresses the failure modes identified in this audit. MassSpecGym v1.5 is publicly available at https://github.com/pluskal-lab/MassSpecGym. --- ### SEAGAN: domain-Specific and Edge-Aware Graph Attention Network for Dynamic Plant Processes Authors: Antriksh Srivastava, Soumyashree Kar Published: 2026-06-17 Categories: cs.LG arXiv: https://arxiv.org/abs/2606.19623v2 Graph neural networks (GNNs) offer a flexible framework for learning from scientific data with physical, biological, or functional associations. One promising domain is plant physiology, where observed responses result from several interacting processes that are difficult to isolate, even with human intervention. A key example is the A-Ci curve, which relates the net CO2 assimilation rate (Anet) to leaf intercellular CO2 concentration (Ci) and is also used to estimate photosynthetic parameters in biophysical models. However, accurate estimation requires accurate identification of the active biochemical limiting state at each curve point, which is a major source of uncertainty. Here, we express the limitation-state identification in A-Ci curves as a graph-based node classification problem. A graph representation of the A-Ci curve is created using distance-based k-nearest-neighbor (kNN) and auxiliary-signal-guided (ASG) connectivity. The methodology was evaluated against the conventional machine learning baselines, graph-based architectures, and an automated fitting-based benchmark. Results on a large synthetic dataset with known ground-truth limitation states show that graph-based models improve classification, especially near biochemical transition areas. The top-performing configuration, SEAGAN (domain-Specific and Edge-Aware Graph Attention Network for Dynamic Plant Processes), integrates process-aware node features, edge attributes, kNN connectivity, and graph attention with a weighted cross-entropy loss, obtaining an F1-score of 0.857 and accuracy of 0.882. The results suggest that using A-Ci curves as graphs enables better identification of the biochemical limiting condition and reduces the uncertainty associated with both human and automated methods. ---