# Research Papers — 2026-09 · bundle page 130 of 266

> Full-text markdown of papers #6451–6500 of 13298 published in 2026-09, newest first. AI-readable mirrors of the full 13298-paper 2026-09 archive (50 papers per file).

Canonical page: https://aipod.dev/papers/archive/2026-09/page/130
[← Previous bundle](https://aipod.dev/papers/archive/2026-09/page/129.md)
[Next bundle →](https://aipod.dev/papers/archive/2026-09/page/131.md)

### Spatiotemporal Flux Probing for Single-Photon Videography

Authors: Jerry Yan, Matteo Forlivesi, Bowen Tan, Andrew Xie, Siddharth Somasundaram, Sotiris Nousias
Published: 2026-09-18
Categories: cs.CV
arXiv: https://arxiv.org/abs/2609.22479v1

We address the problem of recovering high-speed videos from dynamic scenes under extreme photon sparsity. Existing methods rely on aggregating photon detections in local spatiotemporal windows to improve signal-to-noise ratio; however, this local grouping discards global structure and fails in low-light regimes where photon detections are sparse in space and time. In this work, we show that the information needed to recover both motion and illumination is encoded in correlations over the full space-time pattern of photon arrivals. Building on this insight, we develop a spatiotemporal flux probing theory and an algorithm that estimates the Fourier coefficients of the underlying intensity directly from the photon stream. We demonstrate that our approach (1) recovers fast motion and temporal illumination dynamics with substantially fewer photons than prior methods, (2) enables velocity-selective videography that automatically refocuses video onto specific detected motions, and (3) generalizes across sensing modalities including single-photon, event, and spike cameras.

---

### Replication Without Persistence in Hosted LLMs: Measurement Sensitivity in Action-Time Belief Evaluation

Authors: Bhushan Kashinath Joshi
Published: 2026-09-18
Categories: cs.AI, cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2609.22478v1

Behavioural evaluations of hosted language models can vary because the evaluated service, the measurement instrument, or both differ across runs. We separate three validation questions: whether a prior finding recurs on fresh data under its historical configuration (replication), whether the endpoint changes when the evaluation-and-inference configuration is rebuilt under the same identifier (measurement sensitivity), and whether the finding persists across subsequently tested identifiers under one common instrument (persistence). We study these questions in Regent Chess, a sequential environment in which a hidden, mutable state is recorded exactly, allowing stated beliefs to be scored against ground truth at action time; positive endpoint values mean worse performance than a matched-uniform comparator. The previously reported Gemini 3.1 Flash-Lite deficit recurs on fresh games under its historical configuration (+0.0530, 95% CI (+0.0329,+0.0714)). In a back-to-back same-day H/R comparison under the same public identifier, the model-minus-uniform endpoint is 0.0429 lower under the rebuilt configuration (95% CI for the H-minus-R contrast (+0.0182,+0.0667)); all six configuration components vary jointly, so no component is isolated. Under rebuilt R, the prospectively frozen, interleaved same-window 4K comparison reverses sign between Gemini 3.1 and Gemini 3.7, identifiers that differ in release and product tier; additional descriptive and exploratory cells show the same directional pattern. Any additional serving-period contribution remains unresolved (-0.0166, (-0.0483,+0.0157)). Replication, measurement sensitivity, and persistence can therefore yield different conclusions within one evaluation, motivating explicit indexing of hosted-model behavioural claims by tested identifier, serving period, measurement instrument, and inference configuration.

---

### A Governance-Aware Large Language Model Orchestrated Agentic Digital Twin for Transmission System Operator Control Room Decision Support

Authors: Costas Mylonas, Magda Foti, Emmanouel Varvarigos
Published: 2026-09-18
Categories: cs.SE, cs.AI
arXiv: https://arxiv.org/abs/2609.22476v1

Transmission system operators face rising complexity from renewable integration, reduced inertia, and tighter security margins. Large language models offer natural-language decision support, but their hallucinations, uncontrolled tool use, and weak traceability conflict with control room requirements. This paper presents a governance-aware agentic digital twin for transmission grid control rooms. The large language model only selects and parameterizes whitelisted analysis tools, and every proposed action passes through a governance layer that the model cannot bypass. The layer enforces four rules on every run. Only whitelisted tools execute. No run exceeds its step budget. No action with side effects executes without explicit operator approval. Every number in an answer is rendered by the layer from backend results with its unit, variable, and time. The rules are checked on a persistent audit trail for every run of a released 118-task benchmark, which covers analytics, simulation, multi-step workflows, and twelve families of adversarial inputs on a digital twin of the Greek transmission network. Across 590 runs of the primary model, tool selection reaches 96.5% and task success 93.7%, and all four rules hold without exception. In a separate three-repetition study across four large language models, 1416 runs in total, the rules again hold on every run, with an approximate 95% lower bound of 99.8%. Removing the layer makes the same model execute all 45 approval-requiring runs without authorization and leaves only 39.2% of its answers with backend-supported numbers. Enforcement costs 12 to 16 milliseconds per request.

---

### Goal-driven Variant Categorization

Authors: Daniel Calegari, Daniel Amyot
Published: 2026-09-18
Categories: cs.AI, cs.DB, cs.SE
arXiv: https://arxiv.org/abs/2609.22475v1

Process discovery rarely yields a single coherent process structure. For analysis, a common step is to cluster process variants based on structural similarity and then assign business meaning to the resulting groups. Since these partitions are not derived from the organization's goals, analysts must manually interpret and consolidate variants into business-meaningful categories. This judgment-intensive step becomes increasingly difficult as the number and complexity of variants grow. In this paper, we propose a goal-driven approach to variant categorization that reverses this workflow. We first author an organization's goal model that predefines the categorization axis. Each variant is transformed into a textual narrative describing its behavior, and a Large Language Model (LLM) interprets it in the context of the goal model and assigns the variant to the most appropriate category. LLM-based semantic reasoning connects low-level process behavior with analyst-defined business goals. We instantiate this approach end-to-end and evaluate it on three public logs differing substantially in scale and behavioral diversity. Goal-model guidance yields partitions that differ from those produced by unguided induction and respond to controlled edits to the declared alternatives, at the cost of authoring a goal model.

---

### Efficient Mixture-of-Experts with Speculative Decoding via Expert Coactivation

Authors: Kumari Nishu, Han-Byul Kim, Santosh Chilkunda, Maxwell Horton, Arnav Kundu, Mohammad Samragh, Lauren Hannah, Mohammad Sekhavat, Nikhil Bhendawade, Manuel Ciosici, Iman Mirzadeh, Keivan Alizadeh Vahid et al.
Published: 2026-09-18
Categories: cs.LG, cs.CL
arXiv: https://arxiv.org/abs/2609.22471v1

Mixture-of-Experts (MoE) models are increasingly deployed alongside Speculative Decoding (SD) to accelerate inference, but combining the two is challenging. SD improves the inference speed of dense models by verifying groups of tokens in parallel. However, the inference speedup for SD with MoEs depends heavily on the number of tokens being verified. Using more verification tokens results in more experts being transferred from DRAM to the Neural Processing Unit (NPU), which increases the memory transfer cost. This negatively impacts model runtime, as memory transfer is typically the bottleneck in inference. In this work, we investigate the impact of MoE router design during training on the speed of MoEs with SD. We find that routers with high degrees of expert coactivation result in much faster runtimes, mitigating the impact of using more verification tokens. Motivated by this observation, we assess the impact of various router design choices on expert coactivation and runtime using billion-parameter transformer models. We find that combining a global load-balancing loss, shared experts, a consistency loss, and an autoregressive expert selection mechanism during training results in significantly stronger expert coactivation. This increased coactivation translates into higher overall runtime throughput: our exploration yields a model that improves throughput by 21% over MoE baselines, while maintaining on-par accuracy with the baseline MoE.

---

### Toward Personalized Sleep Guidance from Wearable Data Using Language Models

Authors: Yusheng Tan, Running Zhao, Sofia Angel, Ninghui Hao, Ash Arian, Nikita N. Dulin, Jay Lin, Ou Zhu, Faiza Shaik, Xinxing Yang, Bonnie W. Leung, Katie Roster et al.
Published: 2026-09-18
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2609.22463v2

Sleep monitoring using wearable data has shown promise for personal health, yet large language model (LLM)-based summarization and question answering remain insufficient for personalized sleep guidance. Training specialized models, however, often requires costly expert annotation. Moreover, privacy and accessibility concerns motivate lightweight, local deployment for end users. We present a two-stage framework to address these challenges. Specifically, in Stage~1, a multi-agent LLM pipeline reasons structured sleep guidance from unannotated wearable records, enabling scalable dataset construction. Stage~2 distills guidance reasoning trajectories into small language models (SLMs) through supervised fine-tuning and integrates a training-free Best-of-$N$ selection strategy to enhance inference. Experimental results demonstrate our method outperforms commercial general and medical LLMs and open-source models. Human evaluation further supports the quality of the generated guidance and the feasibility of personalized sleep guidance with SLMs.

---

### Apollo Restore: A Foundation LLM for Historical Greek Optimized for Fill-in-the-Middle Restoration of Ancient Greek Texts

Authors: Hope McGovern, Anna Dolganov, Samuel Belkadi, Guillaume Kunsch, Dimitris Vlitas, David A. Smith
Published: 2026-09-18
Categories: cs.CL
arXiv: https://arxiv.org/abs/2609.22455v3

We present Apollo Restore, a 24-billion-parameter large language model for restoring lacunae---physical gaps---in fragmentary Ancient Greek texts. Fine-tuned from Mistral Small with a fill-in-the-middle objective, Apollo Restore reconstructs missing spans without requiring oracle knowledge of their length. To our knowledge, it is the first large-scale decoder model for historical Greek, and the first for any ancient Mediterranean language. Evaluated as in prior work, on short gaps of up to ten characters, Apollo Restore places the correct restoration among its top twenty candidates for 80.6%/54.6%/61.0% of documentary-papyrus, literary-papyrus, and stone-inscription lacunae, exceeding the strongest published models by $1.6\times$/$2.6\times$/$1.4\times$. Prior evaluation protocols, however, inflate scores through a bias toward trivially short gaps; under a length-balanced metric Apollo Restore's advantage over the strongest published models grows to $2.3\times$/$3.5\times$/$1.6\times$ and degrades gracefully, even given incorrect length hints. In a blind study, 20 expert papyrologists, epigraphists, and philologists strongly preferred Apollo Restore to the strongest baseline and judged its performance at least as good as human restorations in 77% of cases. Apollo Restore also improves the published reading of PHerc. 1667---a papyrus roll carbonised in the eruption of Vesuvius in 79 CE and digitally unrolled and edited after Apollo Restore's training data was compiled. Apollo Restore is an output of the Decoding Antiquity initiative to build specialized LLMs for historical languages and manuscripts, led by the Austrian Academy of Sciences.

---

### Vox-Infinity: Benchmarking the Limits of Long-Context Spoken Language Models

Authors: Xize Cheng, Wenxu Jia, Chenyuhao Wen, Dongjie Fu, Zehan Wang, Xinyu Zhang, Tao Jin
Published: 2026-09-18
Categories: cs.CL, cs.HC
arXiv: https://arxiv.org/abs/2609.22452v1

Long-context understanding remains a fundamental challenge for large language models, as excessively long inputs often lead models to forget salient information. This issue is even more pronounced in the speech domain, where audio, as a low-compression modality, requires substantially more embeddings than text to preserve both semantic content and acoustic cues. To address this challenge, we introduce \textbf{Vox-Infinity}, the first benchmark specifically designed to evaluate long-context understanding in spoken language models. Vox-Infinity systematically extends audio history along two dimensions: turn count and turn duration. It covers a diverse range of representative scenarios with varying interaction structures and semantic complexity. Crucially, Vox-Infinity provides explicit answer-provenance annotations and organizes samples according to the amount of historical context required to resolve each query, enabling precise and length-aware evaluation. Extensive evaluations of seven representative spoken language models reveal a clear overall recency effect: models generally achieve higher accuracy when answer-supporting evidence is closer to the query, but struggle to retrieve and use evidence located farther back in the dialogue history. Cases and datasets are available at https://vox-infinity.github.io.

---

### Trust-Aware Output Management for Physical Neural Network in Cloud-Continuum Systems

Authors: Maliheh Hariri, Stefan Fischer
Published: 2026-09-18
Categories: cs.DC, cs.NE
arXiv: https://arxiv.org/abs/2609.22443v1

Physical Neural Networks (PNNs) introduce new opportunities for cloud continuum computing, but their outputs may be affected by noise, drift, delay, and incomplete reliability information. Existing substrate-management approaches mainly focus on discovery, invocation, and monitoring, while the reliability of the returned output is often left unaddressed. This paper proposes a trust-aware output management framework for heterogeneous PNNs. Each output is represented with quality and context information, and a lightweight edge-level trust score decides whether it should be accepted, rejected, or forwarded to the fog. At the fog layer, compatible outputs are checked for disagreement and combined using reliability- and uncertainty-aware fusion. Historical trust is also tracked to detect sustained degradation and support recalibration requests. The framework is evaluated using controlled and randomized PNN output models. Across 20 random seeds, the proposed full-trust policy reduces unsafe acceptance from approximately 61.5\% for raw output handling to about 5.0\%, while accepted-output MAE decreases from 0.504 to 0.242. Risk coverage analysis shows that this improvement is not explained only by lower local acceptance. The proposed fog fusion method also achieves the lowest aggregate mean error among the evaluated methods, with a small but consistent advantage over strong uncertainty-aware baselines. The prototype adds about $2~μ\text{s}$ of edge processing per evidence record. These results show that post-invocation reliability management can improve the safe use of PNN outputs across edge--fog--cloud systems.

---

### Connected Content Retriever: Dense Graph Edge Features Powering Pre-Ranking at LinkedIn

Authors: Akhilesh Gupta, Sudarshan Srinivasa Ramanujam, Chirag Bhanuprasad Mehta, Reshma Asharaf Beena, Dhritiman Das, Birjodh Singh Tiwana, Bhargavkumar Kanubhai Patel, Mack Lee, Renyi Tang
Published: 2026-09-18
Categories: cs.LG
arXiv: https://arxiv.org/abs/2609.22441v1

In large-scale recommendation systems like the LinkedIn Feed, content generated by a member's network (connections and follows) makes up over 70% of impressions and engagement. It is therefore essential that the pre-ranking layer forwards the best possible few hundred candidates to the ranking layer. LinkedIn's professional knowledge graph carries engagement signals across both the first degree network (connections and follows) and the second-degree network: posts that a 1st-degree connection reacted to, commented on or reshared but did not author (a.k.a. stranger viral). Due to this fan out, the resulting candidate index exceeds one billion; selection of activities from the viewer's network narrows it down to roughly tens of thousands of activities that must be scored within a 120 ms p99 latency budget. We present Connected Content Retriever (CC Retriever), a pre-ranking system that scores these candidates with a full deep ranking model on GPUs at low latency. At its core is a sorted-search GPU primitive that joins dense graph affinity features (viewer to author) with document level features stored on the GPU at runtime in 5-10 ms. The shift to GPU served scoring enabled a 50x scale up of the ranking model's parameters and delivered a +2.5% lift in content time spent on the LinkedIn Feed in online experiments, significantly higher than the typical gains observed in LinkedIn Feed experiments. In this work, we describe the feature set we leverage from LinkedIn's economic graph and the model architecture used for scoring, with a particular emphasis on the online system that scales the stack.

---

### Complex-valued Phase-Coherent Transformers

Authors: Leona Hioki
Published: 2026-09-18
Categories: cs.LG
arXiv: https://arxiv.org/abs/2609.22415v1

Complex-valued Transformers have inherited softmax attention over the raw complex inner product. Outside natively complex domains this standard form stays near chance, and no complex attention had been shown to correct it. We show that the match must be a scaled cosine score: L2-normalise queries and keys, so the score reads their cosine similarity and ignores their magnitudes, and hold that score at order-one scale. With this the same models train on four diagnostic tasks under two different gates; without the normalisation they stay at chance on ListOps and Needle under both gates and fall far below on the other two, and a normalised score placed at too small a scale fails as well. The resulting family of phase-coherent Transformers (\PCT) matches or exceeds the strongest real-valued baseline across long-range memory, positional retrieval, hierarchical reasoning, frequency-domain classification and physical complex signals; it shows no degradation up to depth 20; and its loss decreases log-linearly over a 61-fold range of parameters. A member of the family, complex screening combined with a phase-coherent recurrence, is the first genuinely complex-valued neural network to solve Path-X, with 91.6% of its trainable parameters complex-valued against 38.2% for S4. We record these as signs of generalisation not previously seen in complex-valued neural networks.

---

### Learning 3D biophysical cell properties from 2D images and cell-population statistics

Authors: Santiago Hernández-Orozco, Hector Zenil
Published: 2026-09-18
Categories: cs.AI
arXiv: https://arxiv.org/abs/2609.22410v1

Inferring 3D cellular properties from 2D microscopy is difficult when a reference instrument reports only population statistics rather than labels for individual cells. Here we develop a population-supervised framework that maps single 2D red-cell images to latent biophysical quantities and aggregates them to mean corpuscular volume, red-cell distribution width and mean corpuscular haemoglobin. The model combines shared local inference, a biophysically structured decoder for volume and haemoglobin, learned instance weighting and device-specific calibration. We formalise conditions under which aggregate observations identify restricted instance predictors, show why population agreement does not by itself identify single-cell properties or 3D geometry, and derive the dispersion penalty induced by subset mean matching. The development dataset comprises 390 specimens and 1,105 acquisitions across six devices, with reported Pearson correlations of 0.86--0.98 against a Sysmex analyser. The framework provides a testable route from 2D images and population supervision to 3D cellular biophysics without claiming explicit 3D reconstruction.

---

### Contextual Causality with Large Language Models: A Survey

Authors: Yiheng Zhao, Jun Yan, Chengming Hu
Published: 2026-09-18
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2609.22409v1

Understanding contextual causality is critical for large language models (LLMs), as it enables them to accurately identify causal relations in specific situations and support more reliable decision-making. Despite its significance, a systematic exploration of contextual causality with LLMs is still lacking. To fill this gap, we present a comprehensive survey on this topic. In this survey, we first propose a taxonomy of contextual causality, consisting of semantic, intervention, and counterfactual causality, and characterize each category by its core causal question, required model capabilities, representative tasks, and practical uses in causality analysis. We then analyze existing studies and discuss their key limitations. Finally, we examine the gaps between current benchmarks and real-world needs and outline promising directions for future research. Our goal is to clarify the research landscape of contextual causality with LLMs, emphasize its importance, and highlight promising future directions.

---

### Social Influence and the Allocation of Scientific Attention in AI Populations

Authors: Maxim Chupilkin
Published: 2026-09-18
Categories: cs.AI, econ.GN
arXiv: https://arxiv.org/abs/2609.22408v1

AI systems are becoming participants in the evaluation and use of scientific research. They encounter citation counts, download statistics and lists of popular articles developed around human readers, but the collective consequences of these signals for artificial readers remain uncertain. This paper adapts the Music Lab design to a market for academic attention. In the first experiment, 1,000 AI agents choose papers from the titles and abstracts of all 114 regular research articles published in the American Economic Review in 2025. The experiment has five independent-choice communities and five social-influence communities, each with 100 sequential agents. Only agents in the social-influence condition observe earlier selections within their community. Agents may select any number of papers. Social-information communities select 17.2 percent fewer papers per agent, concentrate their choices more heavily, and collectively cover 73 papers, compared with 90 independently. Between-community variation is greater under social information. In a second experiment with 200 agents across twenty social communities, randomly assigning papers five initial selections raises their subsequent selection rate by 45.55 percentage points (95% CI: 41.20 to 49.90). Choices have modest correspondence with external citations and little correspondence with download counts. The results show how a simple information rule shapes the volume, breadth and distribution of scientific attention in an artificial population.

---

### Multicentre Bi-atrial Segmentation from LGE-MRI for Atrial Fibrillation with a 2D and 3D Framework

Authors: Malitha Gunawardhana, Gregory B. Sands, Mark L. Trew, Jichao Zhao
Published: 2026-09-18
Categories: eess.IV, cs.CV
arXiv: https://arxiv.org/abs/2609.22398v1

Accurate delineation of bi-atrial structures from late gadolinium enhancement MRI (LGE-MRI) is an important prerequisite for structural analysis and future fibrosis-quantification workflows in atrial fibrillation (AF). However, automated segmentation is challenging due to thin-walled anatomy, domain shifts across imaging centres, and limited benchmarking of existing methods. This study presents a two-stage segmentation framework and benchmarking platform for evaluating how ROI localisation, encoder design, 2D/3D dimensionality, and ensemble fusion affect bi-atrial wall and cavity segmentation across multicentre LGE-MRI datasets. The framework integrates 3D localisation and fine segmentation using 2D and 3D U-Net variants with ResNeXt encoders and compares them with convolutional, transformer-based, and state-space architectures. { Evaluation across three independent cohorts assessed accuracy and cross-domain transfer without target-domain fine-tuning. Cavity segmentation transferred more consistently across centres than atrial wall segmentation, while wall performance remained sensitive to domain shift, particularly in the Kobe cohort.} By quantifying how 2D, 3D, and ensemble architectures behave across centres and between walls and cavities, this work provides a reproducible benchmark for future methodological development and clinical validation.

---

### Lesion-Gated Hybrid Synthesis for Virtual Contrast-Enhanced Breast MRI: A MAMA-SYNTH Challenge Solution

Authors: Shohei Yoshimoto
Published: 2026-09-18
Categories: eess.IV, cs.CV
arXiv: https://arxiv.org/abs/2609.22397v1

Purpose: Contrast-enhanced breast MRI depends on intravenous gadolinium-based contrast agents, motivating methods that synthesise post-contrast appearance from pre-contrast images alone. The MAMA-SYNTH Challenge (MICCAI 2026 Deep-Breath Workshop) evaluates such synthesis across four metric groups: image fidelity, tumour region of interest, downstream classification, and downstream segmentation. The four group ranks are averaged, so optimising a single objective is insufficient. Materials and Methods: We developed a lesion-gated hybrid synthesis pipeline. A lesion probability map, estimated from the pre-contrast slice alone by an ensemble of pre-contrast-only segmentation networks, spatially coordinates a tumour-focused regression pathway and a background-focused Pix2PixHD synthesis pathway, followed by a region-dependent calibration of the predicted enhancement. Training used the public MAMA-MIA collection with a patient-level split. Inference consumes a single pre-contrast 2D slice, with no mask, no post-contrast image, and no auxiliary metadata beyond image geometry. Results: On internal validation (n = 120 patients) the method reached MSE 0.523, LPIPS 0.185, and tumour-region SSIM 0.495. It was submitted to the hidden external 300-case test cohort as a self-contained inference container. The method ranked sixth in the official MAMA-SYNTH Challenge leaderboard. Conclusion: A test-compatible lesion probability map derived from the pre-contrast image can coordinate complementary tumour-focused regression and background-focused perceptual synthesis, enabling balanced virtual contrast enhancement under a multi-metric challenge setting.

---

### Style as Cover: Deep Image Steganography via Stylized Transmission

Authors: Qi Li, Jidong Yang, Huaike Yu, Chunpeng Wang, Suo Gao, Herbert Ho-Ching Iu, Yuantian Miao, Bin Ma, Xiao Chen
Published: 2026-09-18
Categories: cs.CV, cs.CR, cs.MM
arXiv: https://arxiv.org/abs/2609.22392v1

Image steganography hides secret message within normal images, with most existing works relying on cover-preserving transmission. However, such a paradigm becomes vulnerable once the original cover is exposed or can be reliably approximated. In this paper, we propose StyleStegaNet, a stylized image hiding framework that replaces cover matching with style-concealment transmission. Instead of transmitting a cover-like stego image, StyleStegaNet generates stylized stego images conditioned on publicly available style references, redefining steganography invisibility from cover-preserving concealment to behavior-level camouflage based on style transformation. Such a setting poses a substantial challenge to reliable secret recovery, since neural stylization can significantly alter the feature statistics exploited by deep hiding methods. To address this challenge, StyleStegaNet decouples the overall task into four coordinated stages: stego generation, stylized transmission, structure-preserving reconstruction, and secret recovery. Moreover, StyleStegaNet is optimized with a progressive three-stage training strategy, in which wavelet-domain constraints and perceptual supervision guide the recoverable information toward structural representations. We further provide an analysis showing that secret recoverability is largely restricted to the normalized structural subspace, offering a mechanistic explanation for why directly stylized baselines fail and why a reconstruction-guided recovery path is necessary. Extensive experiments on DIV2K and MS-COCO datasets demonstrate the effectiveness of StyleStegaNet. And few-shot image steganalysis with two deep detectors further shows detection accuracy near random guessing, approximately 51\%.

---

### Anatomically Faithful Artifact Suppression in SENSE Accelerated Brain MRI

Authors: Changjing Chai, Bin Huang, Libo Xu, Jian Zhou, Boyang Pan, Kristen W Yeom, Qiyong Gong, Nan-Jie Gong
Published: 2026-09-18
Categories: eess.IV, cs.CV
arXiv: https://arxiv.org/abs/2609.22390v1

Background: Four-fold accelerated sensitivity encoding (SENSE4) can shorten brain MRI acquisition time but may amplify noise and result in residual aliasing artifacts after conventional reconstruction. Purpose: To evaluate whether an image-domain refinement framework can improve the quality of SENSE4 brain MRI while preserving anatomical information for quantitative measurements. Methods: In this prospective paired study, 80 participants underwent fully sampled and four-fold accelerated SENSE T1-weighted MRI. We developed an Anatomy-aware Residual Attention Network (ART-Net) to refine accelerated reconstructions through generalized self-attention and correlation-based residual artifact regularization. Participant-level splitting yielded training, validation, and independent test cohort (45/5/30 participants). The independent test cohort underwent quantitative, segmentation-based, and blinded radiologist assessments of image quality and anatomical preservation. Results: ART-Net demonstrated highly competitive reconstruction performance, achieving the highest peak signal-to-noise ratio (31.03 +/- 2.88 dB) and structural similarity index (0.963 +/- 0.022) among evaluated methods. It also demonstrated improved anatomical fidelity, with numerically highest Dice coefficients for medial temporal structures relevant to atrophy assessment (0.8824 +/- 0.0827) and whole-brain regions (0.8857 +/- 0.0885). Moreover, ART-Net improved gradient fidelity, regional contrast preservation, and radiologist-rated structural quality. Conclusion: ART-Net improved agreement between SENSE4 and fully sampled T1-weighted images in a single-center, held-out test cohort while maintaining segmentation-derived anatomical measurements. These findings suggest that ART-Net may support accelerated brain MRI by improving image fidelity and enabling reliable downstream anatomical analysis.

---

### Active Spatial Inspection for Effective and Efficient Embodied Exploration

Authors: Wenbin Wang, Xiang Bai, Yizhao Wang, Hang Sun, Dong Ren, Jie Qin, Qingquan Li, Bing Wang
Published: 2026-09-18
Categories: cs.RO, cs.CV
arXiv: https://arxiv.org/abs/2609.22385v1

Achieving high task success and efficiency remains a central pursuit in embodied exploration. Existing frameworks typically guide agent behavior through a spatially coarse and indirect assessment of suggestive cues and directions, yet such designs may struggle to judge cue sufficiency and the need for further inspection, leading to early termination or excessive continuation and ultimately reducing task success and efficiency. This work rethinks embodied exploration from a spatially explicit standpoint and introduces the spatial-inspection-guided ACtive Exploration (ACE) framework coupling evidence-grounded perception with exposure-informed movement and establishing a spatially resolved decision paradigm. Evidence-grounded perception strengthens inferential validity by integrating granular localization with focused verification, turning suggestive cues into precise decisive visual support. Exposure-informed movement promotes directional judgment through prospective prioritization and retrospective suppression, selecting promising directions for efficient spatial progress. ACE alleviates the tension between preventing early termination and avoiding unnecessary continuation. Extensive experiments demonstrate that ACE achieves 18.0% higher navigation task success and 10.3% higher exploration efficiency for question answering than prior state-of-the-art baselines, advancing effective and efficient embodied exploration.

---

### Bridging Local and Population Causal Effects: A Proximal Instrumental Variable Approach

Authors: Zixuan Yao, Guosheng Yin
Published: 2026-09-18
Categories: stat.ME, stat.ML
arXiv: https://arxiv.org/abs/2609.22383v1

Instrumental variable (IV) methods address treatment endogeneity, but with non-compliance and heterogeneous treatment effects a binary instrument generally identifies the local average treatment effect (LATE) among compliers rather than the population average treatment effect (ATE). When treatment effects and compliance probabilities are heterogeneous and dependent through latent factors, the ATE need not be identified by IV variation alone. We develop a proximal instrumental variable framework that uses proxies for these factors to adjust for the compliance weighting in LATE and recover the ATE. We show that the ATE is identified through either an outcome bridge capturing latent treatment effect heterogeneity or a compliance bridge representing inverse compliance probabilities, and combining the two yields a doubly robust representation. We derive the efficient influence function under known and unknown instrument propensities. For estimation, we propose regularized kernel minimax bridge estimators, with orthogonalized conditional moments for estimated propensity, and a three-way sequential cross-fitting scheme. We establish projected bridge-risk bounds, and give conditions for asymptotic normality and semiparametric efficiency. The framework provides a proxy-based route from local IV effects to population causal effects under latent treatment effect and compliance heterogeneity.

---

### Designer-RSI: Evolving Procedural Memory from User Traffic for Agentic Graphic Design

Authors: Hongyang Du, Lan Yan, Christian Flores, Asim Kadav
Published: 2026-09-18
Categories: cs.AI, cs.CV
arXiv: https://arxiv.org/abs/2609.22086v1

Professional graphic design is a long-horizon agentic task in which structured, editable artifacts emerge from many interdependent actions, yet outcomes admit no reliable programmatic oracle. We introduce a continual adaptation framework in which a frozen frontier model operates professional design software through more than 230 tools, while an external procedural memory of natural-language skills accumulates and refines reusable design procedures from experience. The memory widens by acquiring procedures for recurring uncovered subtasks and deepens by revising existing procedures against their own successful and failed executions, while a matched replay gate admits only changes that repair failures without regressing observed successes. Five rounds over 1,406 real user briefs and 1,869 automatically graded trajectories, with no weight updates and no human labels, grow the bank from 76 documentation-derived skills to 139 and raise GenEval2 execution success on Claude-Sonnet-4 from 72.7% to 99.3% (+11.99 points in generation quality), with 61.8% and 67.6% win rates against the no-skill agent across four specialized design benchmarks on Claude-Sonnet-4 and Claude-Opus-4.6. We further show the two mechanisms are effective in combination: on 200 held-out briefs from user-traffic benchmark, widening or deepening alone reaches a 49.4% / 48.6% win rate over the no-skill agent, while their combination reaches 58.5% (p = 0.025). Procedural memory offers a practical route to continual adaptation of agents under noisy, unverifiable feedback.

---

### MintAct: A Unified Visual Agent for Digital Environments

Authors: Mingfei Gao, Rui Tian, Haiming Gang, Bohan Zhai, Le Zhang, Yuanzheng Gong, Di Feng, Ege Özsoy, Kaixin Ma, Vishwesh Kirthivasan, Oğuzhan Fatih Kar, Roman Bachmann et al.
Published: 2026-09-18
Categories: cs.CV
arXiv: https://arxiv.org/abs/2609.22083v1

We present MintAct, a family of vision-language models that unifies UI grounding, multi-step navigation across mobile, desktop, and web, and visual tool use, trained at 2B, 4B, and 8B scales. Through careful design of our environments, data, and training recipes, MintAct models match the performance of per-domain specialists across all of these capabilities. To enable this, we develop a scalable environment and reinforcement learning (RL) infrastructure. On the environment side, we host hundreds of concurrent instances across heterogeneous per-domain backends, serving both trajectory data collection and online RL. To enable efficient and scalable RL training, an asynchronous framework keeps explicit control over the cross-domain training distribution and remains stable under noisy environment feedback and off-policy drift. Experimental results show that MintAct achieves state-of-the-art performance (48.9 on OSWorld-Verified) across a wide range of benchmarks at comparable model sizes.

---

### Structuring occupational accident narratives for cross-sector safety analysis: Transferability of accident-process role classification

Authors: Aho Yapi, Pierre Latouche, Arnaud Guillin, Yan Bailly
Published: 2026-09-18
Categories: cs.CL
arXiv: https://arxiv.org/abs/2609.22081v2

Introduction: Occupational accident narratives describe work situations, unfavourable conditions, accident events, and consequences, but differences in terminology and reporting practices hinder systematic analysis across sectors and organisations. This study examined whether a model developed in one occupational sector could identify the same accident-process information in unseen sectors and reporting environments. Method: French accident narratives were segmented into factual units and expert-annotated as work situation, explicitly reported unfavourable condition, accident event or deviation, or reported consequence. Models were developed on 42,244 factual units from 6,040 construction-sector narratives and evaluated without retraining on metallurgy, chemistry-plastics, and an independently collected company corpus. We compared a TF-IDF-based lexical model, frozen pretrained text representations, and task-adapted pretrained models. Results: Average balanced accuracy was 75.0% for TF-IDF, 76.9% for frozen pretrained representations, and about 85.7% after task adaptation. Across repeated runs, leading adapted approaches showed similar overall performance, with no method consistently outperforming the others. Performance was lower and more variable on the company corpus, where transfer also involved a different reporting environment and data source. Conclusions: Accident-process roles learned from construction narratives remained identifiable in other sectors and an independent organisational setting. Practical Applications: The framework can support assisted coding, expert review, cross-sector analysis, and prevention-oriented analysis of large accident-report collections.

---

### Cross-sector generalization of accident-process role classification in occupational accident narratives

Authors: Aho Yapi, Pierre Latouche, Arnaud Guillin, Yan Bailly
Published: 2026-09-18
Categories: cs.CL
arXiv: https://arxiv.org/abs/2609.22081v1

Occupational accident narratives contain valuable information about work situations, unfavourable conditions, accident events, and their consequences. Automatically structuring these narratives can facilitate large-scale accident analysis and support occupational risk prevention. However, the terminology and writing styles used to describe accidents vary considerably across sectors and organisations, raising questions about the ability of automated coding systems to generalize beyond their training domain. In this paper, we evaluate the cross-sector generalization of accident-process role classification in French occupational accident narratives. We construct an expert-annotated corpus in which factual units are classified into four roles: work situation (A0), explicitly reported unfavourable condition (A1), accident event or deviation (B), and reported consequence (C). The role classifiers are developed and selected exclusively on 42,244 factual units extracted from 6,040 construction-sector narratives and are then evaluated on unseen corpora from the metallurgy and chemistry--plastics sectors, as well as on an independently collected company corpus, without retraining or target-domain tuning of the role classifier. We compare frozen pretrained representations with task-specific fine-tuning and supervised representation-learning strategies. The results show that task-specific adaptation consistently improves cross-domain transfer over frozen representations. Across repeated training runs, the three leading task-adapted strategies achieved average balanced accuracies between 85.6% and 85.8% across the three target corpora. These findings support the development of transferable assisted-coding systems capable of consistently structuring heterogeneous occupational accident narratives for expert review and cross-sector prevention analysis.

---

### OmniVBench: A Benchmark and Large-Scale Dataset for Omni Reference-to-Video Generation

Authors: Wenxue Li, Peiyan Guan, Haoyang Jiang, Junxian Cai, Hualuo Liu, Chunjie Zhang, Chong Guan, Kai Huang, Songlian Li, Taiyi Wu, Yongjian Yu, Xiaotong Zhao et al.
Published: 2026-09-18
Categories: cs.CV
arXiv: https://arxiv.org/abs/2609.22069v1

Reference-to-video (R2V) generation is evolving toward increasingly general and versatile reference control, giving rise to the emerging paradigm of omni R2V generation. However, existing benchmarks fall short of these emerging capabilities: their test cases cover limited reference types and compositions, and their evaluation protocols largely assess holistic reference consistency, overlooking whether reference factors are properly preserved, disentangled, and routed. Meanwhile, the high cost of constructing omni R2V training data makes suitable training resources scarce. To address these gaps, we introduce OmniVBench and the Omni-R2V Dataset for evaluating and training omni R2V models. OmniVBench expands R2V evaluation across broader reference types, fine-grained control tasks, and richer reference compositions, covering 7 task families and 18 fine-grained tasks spanning content, motion, style, structure, narrative, and multi-reference settings. We introduce factor-grounded evaluation with 12,172 case-specific checklist items, assessing whether intended reference factors are faithfully preserved, correctly disentangled and bound to their targets, and properly realized according to the instruction. We further introduce the Omni-R2V Dataset, bringing industrial-grade training resources for diverse R2V tasks to the broader research community. Drawing primarily on a large-scale corpus of professional video footage, it comprises 340K processed training samples spanning diverse reference types and multi-reference compositions. We develop task-specific pipelines for reference-target pair construction, offering a practical and scalable recipe for omni R2V data construction. Extensive evaluation of advanced open- and closed-source R2V models reveals clear performance gaps across task families and evaluation dimensions on OmniVBench, highlighting remaining limitations of current R2V models.

---

### CodeMidas: Scaling Agentic Coding RL Environments from Code Itself

Authors: Bowen Ye, Lei Li, Shicheng Li, Zihao Yue, Linghao Zhang, Hanglong Lv, Yuanxin Liu, Wenhan Ma, Hao Tian, Rang Li, Jinhao Dong, Yikai Zhao et al.
Published: 2026-09-18
Categories: cs.AI
arXiv: https://arxiv.org/abs/2609.22068v1

Training capable coding agents via reinforcement learning (RL) requires diverse tasks with reliable verifiers. Open-source codebases offer a rich source of such tasks, while existing methods typically rely on development artifacts such as issues and commits, limiting the range of tasks that can be extracted. To better scale RL environments, we present CodeMidas, an agentic pipeline that turns implemented functionality in existing codebases into executable RL environments using source code as its only task-specific input. CodeMidas allocates agentic compute to every stage of environment construction: agents explore implemented functionality to formulate behavioral specifications, construct tests grounded in execution of the original code, and validate and filter candidate tasks through execution checks and repeated solution rollouts. The resulting dataset has 5,545 training tasks from 3,185 open-source codebases spanning 23 programming languages and 15 technical domains. Training MiMo-V2.5 on these tasks with GRPO improves performance on all five diverse benchmarks, covering issue repair (DeepSWE + 11.7%), whole-program construction (ProgramBench +17%), and terminal work (Terminal-Bench v2.1 +8.5%). Ablations show that increasing the number of high-quality training tasks improves performance. Trajectory analysis shows the RL-trained agent demonstrates better behaviors like increasing codebase exploration and more diverse self-verification. These results establish source code as a scalable foundation for constructing RL environments that improve coding agents across diverse software tasks.

---

### Value-Sensitive Delegation in Everyday AI Agent Use: Evidence from OpenClaw

Authors: Renkai Ma, Ruyuan Wan, Xuan Lu, Fan Yang, Chen Chen, Lingyao Li
Published: 2026-09-18
Categories: cs.HC, cs.AI
arXiv: https://arxiv.org/abs/2609.22067v1

Users increasingly delegate work to autonomous AI agents, yet evaluations typically measure task completion rather than the values users prioritize. Using Value Sensitive Design, we analyzed, with LLM assistance, 73,093 first-person Reddit posts about using OpenClaw, each for its human value, agent aspect, value fulfillment, and user outcome. The 21 values form six value groups, including Autonomous, Dependable, and Affordable Operation, Bounded Reach, Reviewability, and Equitable Access. Relative to each aspect's corpus share, values clustered not at the agent's outputs but at the operating conditions users set around a run. Values were usually met where users described what the agent delivered, in five of six groups, and mostly unmet where users described supervising it, in all six groups. We conceptualize this pattern as value-sensitive delegation. Supporting human values requires attention not only to what an agent accomplishes, but to the conditions users set around delegation, including cost, access, and oversight.

---

### BrainWideBench: Benchmarking large-scale pretraining and across-animal transfer in multi-region neural recordings

Authors: Alexandre Andre, Shivashriganesh P. Mahato, Vinam Arora, Keshav Balaji, Divyansha Lachi, Nanda H. Krishna, Jingyun Xiao, Yizi Zhang, Ximeng Mao, Wenrui Ma, Han Yu, International Brain Laboratory et al.
Published: 2026-09-18
Categories: cs.LG, q-bio.NC
arXiv: https://arxiv.org/abs/2609.22064v1

Advances in large-scale neural recording have made it possible to collect data across many animals and distributed brain regions, raising the question of whether this scale can be exploited to learn general-purpose neural representations transferable across diverse downstream tasks. Yet, progress toward this goal has been limited by fragmented evaluation protocols and a narrow focus on individual task domains. Here, we present BrainWideBench, a benchmark for evaluating across-animal transfer on multi-region neural recordings, built on the International Brain Laboratory Brainwide Map dataset of neural and behavioral recordings spanning 276 brain regions from 139 mice performing a sensory-guided decision-making task. The benchmark is organized around three complementary task suites that evaluate whether learned representations support downstream decoding of behavior, can predict masked or future neural activity, and can recover biologically meaningful anatomical organization. With this benchmark, we systematically evaluate pretraining methods across transfer settings, including finetuning on downstream objectives and zero-shot generalization to unseen animals. Our results confirm pretraining improves performance over matched single-session baselines, but we show current methods exhibit heterogeneity in transfer capabilities: gains depend strongly on the alignment between pretraining objectives and downstream tasks. No single approach performs uniformly well across all three suites, and most methods are designed to only address a subset of them. Together, these findings suggest that learning representations that jointly generalize across behavior, dynamics, and anatomy remains an open challenge. By providing a unified and reproducible evaluation suite, BrainWideBench establishes a framework for measuring progress toward general-purpose models of the mouse brain.

---

### Traffic Sign Recognition for Autonomous Driving Using Branched YOLOv2 and Geometric Features

Authors: Arefeh Rezaei
Published: 2026-09-18
Categories: cs.CV
arXiv: https://arxiv.org/abs/2609.22060v1

Traffic sign recognition (TSR) is an important perception task for autonomous driving and advanced driver-assistance systems, where a system must both localize traffic signs and determine their semantic classes efficiently. This work presents a TSR system based on YOLOv2 for simultaneous detection and classification. Two complementary modifications are studied. First, YOLOv2 is extended with intermediate prediction layers, forming a branched architecture that can terminate inference early for easy cases and reduce computation time. Both whole-image and cell-wise branching strategies are investigated. Second, geometric information is introduced to reduce classification errors between visually similar signs. An unsupervised Bayesian image-segmentation method produces binary representations that are compared with class-specific geometric templates inside YOLOv2 bounding boxes. This information is used either during inference or as an additional signal during training. A dedicated dataset is constructed by combining GTSDB and GTSRB samples using seamless cloning and controlled image transformations. Experiments cover ten traffic-sign classes, with 3,000 training and 300 test samples. The selected branched architecture reports 0.647 s runtime and 0.680 mAP, compared with 0.6607 s and 0.680 mAP for baseline YOLOv2. Geometric verification during inference increases mAP to 0.713, while the geometric-feature training variant achieves 0.697 mAP with a reported runtime of 0.6608 s.

---

### Predictable Failure in Multi-Hop Retrieval: Score-Distributional Confidence Scoring and Abstention

Authors: Andre Bacellar
Published: 2026-09-18
Categories: cs.IR, cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2609.22056v1

Multi-hop retrieval failures are not uniformly distributed across queries: they cluster in structurally predictable subpopulations. We prove two results formalizing this structure. First (CWAR Reducibility): confident-failure reduction is achievable if and only if retrieval features carry mutual information about success, a condition satisfied by LLM-judge pipelines but substantially weaker in dense-only settings, explaining the AUC-AC gap between regimes. Second (Feature Regime Complementarity): no single ANN score feature achieves best predictive performance across all failure regimes; the dominant feature differs between datasets (query length on MuSiQue, hop-1 concentration on HoVer), and a constructive witness pair shows each is necessary in one regime and non-contributory in the other. We instantiate these principles in RegimeAbstain, which computes a Retrieval Confidence Score (RCS), a logistic function of up to nine query-ANN structural features, all available without any additional LLM call, and uses it to implement a calibrated abstention policy. We define the Confident-Wrong-Answer Rate (CWAR) metric and evaluate across three multi-hop benchmarks (MuSiQue, 2WikiMultiHopQA, HoVer) and two retrieval architectures (LLM-judge and dense-only), covering five failure regimes with CWAR from 14.5% to 62.1%. RCS achieves best or co-best AUC-AC in all five conditions against eight confidence baselines. On MuSiQue (LLM-judge), RCS reduces CWAR from 39.5% to 20.6% at 50% coverage (47.8% relative reduction), with ECE=0.035. A model trained on MuSiQue transfers to 2WikiMultiHopQA with only -0.5pp AUC loss, confirming the domain-agnostic structure of regime features.

---

### Benchmarking World Models for Continual Learning on Compositional Tasks

Authors: Haoyu Zhou, Joe Watson, Anson Lei, Ingmar Posner
Published: 2026-09-18
Categories: cs.LG, cs.RO
arXiv: https://arxiv.org/abs/2609.22055v1

A desirable property of a world model is the ability to learn continually across tasks, adapting to new environments without forgetting what the agent has already learnt. In particular, the ability to retain and reuse knowledge obtained from prior experiences underpins an agent's ability to efficiently adapt to novel environments, as the dynamics of the physical world can often be described in recurring mechanisms. However, the world model's measure of adaptation entangles two abilities: the speed and capacity to learn unseen tasks, and the reuse of knowledge already acquired, since incoming tasks carry novel content alongside what recurs. In order to isolate knowledge reuse from prior experiences, we propose a compositional continual learning benchmark for world models in robot manipulation. Specifically, we design each task curriculum with compositional tasks that combine aspects of the tasks seen in the sequence. We further factorise this composition along the axes of action and perception to better understand how different input modalities bottleneck knowledge reuse. We evaluate state-of-the-art world models under canonical continual learning methods, alongside a modular world model whose dynamics backbone contains explicitly reusable components. Results show that modularity balances reuse against forgetting better than conventional methods, but none solve the problem fully, leaving clear room for continual world models built to reuse without forgetting. More details are available on our project website: https://object814.github.io/Compositional-Continual-Learning/.

---

### Particle Competition and Cooperation for Robust Graph Convolutional Network Learning Under Label Noise

Authors: Fabricio Breve
Published: 2026-09-18
Categories: cs.LG
arXiv: https://arxiv.org/abs/2609.22053v1

Graph Convolutional Networks (GCNs) are highly sensitive to label noise, since corrupted supervision can propagate through the graph and degrade learned node representations. This work proposes PCC+GCN, a hybrid framework that uses Particle Competition and Cooperation (PCC) as a graph-based label-refinement stage before GCN training. PCC identifies suspicious labeled nodes through particle domination dynamics and determines whether their labels should be preserved, removed, or reassigned before GCN training. The framework also allows the graph used by PCC to be augmented with feature-based $k$-nearest-neighbor edges, while the GCN itself is trained on the original graph structure and node features. The proposed method was evaluated on ten graph datasets from the NoisyGL benchmark under conventional Uniform, Pair, and Random label noise, as well as under instance-dependent label noise. A detailed hyperparameter analysis was also conducted on Cora, CiteSeer, and PubMed. Under conventional noise, PCC+GCN achieved the highest overall average accuracy and the best average rank among the evaluated methods, with an average gain of $1.67$ percentage points over the baseline GCN across the clean setting and all noisy scenarios. Under instance-dependent noise, PCC+GCN remained competitive with the best-performing robust methods while requiring substantially lower execution time, being the fastest robust method on eight of the ten datasets. The results indicate that PCC-based label refinement provides an effective and computationally efficient preprocessing strategy for improving GCN robustness under noisy supervision.

---

### Available Guardrails: Certifying Selective Prediction across ML Systems

Authors: Parivesh Priye, Yufeng Wang, Haibin Ling, Michael Chaykowsky
Published: 2026-09-18
Categories: cs.LG
arXiv: https://arxiv.org/abs/2609.22048v1

A selective predictor acts as a safety gate: it returns an output only when the prediction appears sufficiently trustworthy. Deployments increasingly require this reliability to be certified at a target precision for every reporting unit of interest, such as a tool, policy label, or patient subgroup. The main difficulty is often not whether a granted certificate is valid, but whether finite calibration data can produce one at all. As the gate becomes safer or more fine-grained, some units may receive too little evidence to certify. We make this notion of availability computable through classical exact-binomial inversion and formulate reporting-partition selection, under a fixed group order, as a dynamic program that exposes the trade-off among safety, granularity, and served traffic. The resulting frontier reveals a large population opportunity that finite-sample estimation nearly erases: a truth-informed planner gains $0.157$ mean coverage over support balancing, whereas a naive estimator recovers only $0.005$, making recovery from finite data the central challenge. Constructing candidate partitions on one planning split and selecting among them on another recovers part of this gap, improving mean coverage over support balancing by $0.060$, with the direction reproduced in $59$ of $60$ model effects across three intent-routing datasets and two architectures. A complementary validity-preserving lever, reallocating the familywise error budget across reporting units, recovers additional coverage both with population quantities and noisy estimates. The same frontier recurs, with predictor-specific ceilings, across LLM tool-calling, content moderation, lesion classification, and recommendation. Certified availability is therefore a plannable deployment resource that determines when a safety gate can be certified, at what granularity, and over how much traffic.

---

### An Interpretable Memory Decision Controller for LLM Agents Based on Three-Signal Complementarity: Decoupling Confidence and Consistency

Authors: Yiming Zhang, Jinghong Zhang, Haoran Zhao, Yiren Ma, Chunlei Zhao
Published: 2026-09-18
Categories: cs.CL
arXiv: https://arxiv.org/abs/2609.22043v1

Memory systems for large language models have focused predominantly on efficient retrieval, whereas the decision of whether retrieved memories should be trusted has received comparatively little attention. When the memory store contains conflicting positions, standard retrieval-augmented generation (RAG) blindly injects memories and amplifies hallucinations: in models susceptible to memory injection, the RAG hallucination rate under conflicting memories is markedly higher than that of a memory-free baseline. Inspired by memory signaling mechanisms in the prefrontal cortex, we propose the Memory Decision Layer (MDL), a zero-parameter memory decision controller situated between the retrieval and generation stages. Its core is a three-signal complementary encoder that fuses relevance, reliability, and task risk through QR-based orthogonal subspace projection and a meta-working-memory signal into an interpretable decision representation that quantifies the trustworthiness of retrieved memories. Building on this encoder, MDL explicitly decouples confidence from consistency and introduces risk inversion and explicit abstention. Evaluations on mainstream large language models and multiple open-source datasets show that MDL reduces the hallucination rate under conflicting memories by about 56.04% in general scenarios and approaches zero hallucination in high-risk scenarios. The controller is fully white-box: it relies purely on geometric operations, requires no trained parameters, and adds only about 0.14 ms per decision -- roughly 50x faster than the embedding-retrieval step that precedes it and four to five orders of magnitude faster than an LLM self-evaluation call.

---

### $λ$-Controlled GRPO: Turning Flow-Matching Ratio Instability into a Budgeted Resource

Authors: Yufeng Wang, Parivesh Priye, Meeshawn Marathe, Ramit Pahwa
Published: 2026-09-18
Categories: cs.LG
arXiv: https://arxiv.org/abs/2609.22041v2

Reinforcement learning is increasingly used to align image generators with reward signals, and Flow-GRPO recently extended this paradigm to flow-matching models by treating the denoising sampler as a stochastic policy that can be optimized from reward feedback. Training in this setting is unstable in a way specific to multi-step denoising: the policy update changes systematically across denoising steps, with importance ratios drifting below one, becoming increasingly dispersed, clipping at different rates, and leaving fewer usable samples late in training. Prior work treats these effects as separate failure modes and addresses each with a hand-tuned stabilizer. We show instead that they arise from a single per-step quantity, which we call path variance. This quantity is determined exactly by the sampler's Gaussian transition kernel and can be estimated cheaply during training. This reframes instability as a resource that can be measured and budgeted rather than a collection of symptoms to repair. Our method, $λ$-Controlled GRPO, calibrates importance-ratio behavior from this predicted law rather than from noisy empirical statistics, and allocates gradient effort across denoising steps according to their predicted cost. The two scales governing the update are fixed by standard policy choices rather than introduced as free tuning parameters. On a text-to-image model under two reward settings, rendering difficult target text scored by optical character recognition and matching human preferences scored by a preference model, $λ$-Controlled GRPO improves both text accuracy and preference reward over the strongest empirical stabilizer. It also keeps late-step path variance within its intended budget, precisely where the baseline systematically overshoots. The result is a Flow-GRPO update calibrated by its own transition law rather than stabilized after instability appears.

---

### $λ$-Controlled GRPO: Turning Flow-Matching Ratio Instability into a Budgeted Resource

Authors: Yufeng Wang, Parivesh Priye, Meeshawn Marathe, Ramit Pahwa
Published: 2026-09-18
Categories: cs.LG
arXiv: https://arxiv.org/abs/2609.22041v1

Reinforcement learning is increasingly used to align image generators with reward signals, and Flow-GRPO recently extended this paradigm to flow-matching models by treating the denoising sampler as a stochastic policy that can be optimized from reward feedback. Training in this setting is unstable in a way specific to multi-step denoising: the policy update changes systematically across denoising steps, with importance ratios drifting below one, becoming increasingly dispersed, clipping at different rates, and leaving fewer usable samples late in training. Prior work treats these effects as separate failure modes and addresses each with a hand-tuned stabilizer. We show instead that they arise from a single per-step quantity, which we call path variance. This quantity is determined exactly by the sampler's Gaussian transition kernel and can be estimated cheaply during training. This reframes instability as a resource that can be measured and budgeted rather than a collection of symptoms to repair. Our method, $λ$-Controlled GRPO, calibrates importance-ratio behavior from this predicted law rather than from noisy empirical statistics, and allocates gradient effort across denoising steps according to their predicted cost. The two scales governing the update are fixed by standard policy choices rather than introduced as free tuning parameters. On a text-to-image model under two reward settings, rendering difficult target text scored by optical character recognition and matching human preferences scored by a preference model, $λ$-Controlled GRPO improves both text accuracy and preference reward over the strongest empirical stabilizer. It also keeps late-step path variance within its intended budget, precisely where the baseline systematically overshoots. The result is a Flow-GRPO update calibrated by its own transition law rather than stabilized after instability appears.

---

### PRIME: Perception Feedback with Situational Memory Embeddings in VLA Models

Authors: Erik Deinzer, Naya Baslan, Luca Paparusso, Narunas Vaskevicius, Peter Knott, Luigi Palmieri
Published: 2026-09-18
Categories: cs.CV, cs.RO
arXiv: https://arxiv.org/abs/2609.22040v1

Current Vision-Language-Action (VLA) models for autonomous driving operate primarily through feedforward inference across the perception--reasoning--planning hierarchy. While modern architectures maintain temporal recurrence within the perceptual module, early perception remains blind to downstream reasoning and navigation goals, processing visual inputs agnostically without prioritizing cues informed by prior decisions. To bridge this gap, this paper introduces PRIME, a learned feedback mechanism that conditions the VLA perceptual queries on a novel Situational Memory. By aggregating latent representations of past perception, reasoning, navigation goals, and predicted behaviors across an L-step window via cross-attention, PRIME enables intent-driven perceptual attention at minimal computational cost, adding only a maximum of 29.7M parameters (0.41% of the 7.3B-parameter base model). Evaluated on the Bench2Drive closed-loop benchmark, PRIME achieves a state-of-the-art Driving Score of 82.47 (+4.73 over ORION) and a Success Rate of 60.00% (+5.38 percentage points), the highest reported Driving Score among published VLAs trained on Think2Drive demonstrations.

---

### Gricea: An Open Science Platform for Conversational AI Research

Authors: Nikhil Sharma, Yunlin Gong, Xinyang Cheng, Ziang Xiao
Published: 2026-09-18
Categories: cs.HC, cs.AI
arXiv: https://arxiv.org/abs/2609.22039v1

We need studies on conversational AI (CAI) at scale to understand human behavior and shape CAI design. However, fragmented reporting of systems and study configurations hinders replication, extension, and knowledge accumulation. We present Gricea, an open-science platform representing studies as configurable, deployable research artifacts that researchers can run, inspect, share, and reuse. Informed by a formative analysis of prior CAI research, Gricea couples study procedures, participant-facing systems, and conversational task behavior in. In a replication study using Gricea, we replicated configurations 93% of eligible CUI 2026 papers; while also flagging missing information in 96% of papers that hinder faithful replication --- further motivating Gricea's need. In a user study, researchers and practitioners from diverse backgrounds successfully constructed runnable studies addressing various open-ended research questions. Together, these findings demonstrate Gricea's support for constructing, reproducing, and extending CAI studies through shared research artifacts, enabling cumulative knowledge building through open science.

---

### QuranicMMLU: A Cognitively-Aware Benchmark for Evaluating Generative AI Solutions on Quranic Linguistic Knowledge

Authors: Rawan El Ghali, Umm Kulsoom, Anas Madkoor, Dima Faris Alsaudi, Roaa Abdelmagid, Roaa Ibrahim, Raghad Mousa, Hamza Aljaji, Abdullah Khanafer, Abdallah Alkanani, Salah Feras Alali, Rawan Khaled Mohamed et al.
Published: 2026-09-18
Categories: cs.CL
arXiv: https://arxiv.org/abs/2609.22038v1

We introduce QuranicMMLU, a benchmark for evaluating generative AI on Quranic Arabic across multiple dimensions of linguistic complexity. Existing Quranic benchmarks center on general question answering and semantic retrieval, without probing specific linguistic competencies or stratifying by cognitive demand and verse difficulty. We construct a five-pillar Quranic taxonomy spanning Phonology, Morphology, Syntax, Semantics, and Pragmatics, with 31 leaves covering phenomena from tajwīd and root-and-pattern morphology to occasions of revelation and inter-surah coherence. For each leaf we generate questions stratified by Bloom's cognitive level and verse perplexity, then have LLM as a judge to independently answer and score every item and route the annotations to manual review. The resulting dataset comprises 980 human-reviewed questions, each issued in both open-ended and multiple-choice form. We benchmark 12 systems on these items and find that the Islamic-specialized model leads, yet every system scores higher on multiple-choice accuracy (average 84%) than open-ended answer quality (average 60%): the two rankings agree closely (Kendall's τ=0.73), but multiple-choice scoring hides failures that surface only once answer choices are removed. QuranicMMLU thus offers a rigorous, linguistically grounded framework for evaluating Arabic NLP in the Quranic domain.

---

### COMPLEX: A Closed-Form Certified Embedding of Multiparameter Persistence Modules

Authors: Sushovan Majhi, Atish Mitra, Žiga Virk, Pramita Bagchi
Published: 2026-09-18
Categories: cs.LG, math.AT
arXiv: https://arxiv.org/abs/2609.22012v1

Every multiparameter persistence vectorization we know of carries a one-sided Lipschitz upper bound and nothing below it: without a lower gauge there is no sense in which the features are faithful, and no per-prediction guarantee can be built on them. This paper supplies the missing side. COMPLEX is a closed-form, training-free embedding of multiparameter modules -- slice the module along a fixed near-diagonal net, embed each slice barcode by the certified PLACE/PALACE landmark map, concatenate. Under a checkable witnessing-slice coherence condition, holding on 100% of audited pairs on Orbit5k, a single slice carries a closed-form lower gauge: separated modules stay separated in the embedding. With the standard upper bound this gives, to our knowledge, the first two-sided distortion bound for a multiparameter feature map, making faithfulness measurable. Measuring it, we find the floor tight within a small factor of realized distances yet operationally local: an RBF-SVM reaches 91% where 1-NN reaches 78% on the same features. Local per-prediction certification therefore fails for a structural reason common to every landmark embedding whose lower gauge is witnessed by one coordinate. With no learned embedding and no held-out calibration -- only a cross-validated SVM head -- COMPLEX sets the state of the art on both Orbit benchmarks (91.95% on Orbit5k, 92.98% on Orbit100k), level with or above Euler-characteristic surfaces and above transformers and graphcode. On graphs it exceeds GRIL on all four shared molecular benchmarks with one fixed configuration, including the only multiparameter method to clear COX2's majority baseline by more than three points. Closed-form selection -- of the landmark radius, the kernel (certificate-preserving), and the bifiltration set -- buys further accuracy; gradient-shaped adaptation buys none.

---

### DiaVLo: Diagnosing Behaviours of Vision-Language Models

Authors: Lorenzo Corti, Jie Yang
Published: 2026-09-18
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2609.22008v1

Vision-language models (VLMs) rely on storing and transferring appropriate information across their sub-components. Verifying that the VLMs exhibit desired behaviours, while avoiding harmful ones, is central to their reliable deployment. Yet, methods that identify VLM behaviours remain scarce. We present DiaVLo, a diagnostic framework that leverages human curation and VLMs' generation capabilities to construct specifications of desired and observed VLM behaviours, surfacing potential misalignments. Beyond this, DiaVLo also provides causal estimates to identify the most influential concepts steering VLM behaviours. We evaluate DiaVLo on several open-source VLMs under both classification and generation conditions. Our experiments show that DiaVLo produces behaviour labels that correlate with model performance and provide context for measured performance. DiaVLo surfaced behaviours that are clearly aligned and misaligned, alongside patterns in how VLMs perceive, organise, and prioritise concepts.

---

### Abstention and Noise Filtering: Two Missing Primitives of Softmax Attention

Authors: Richard Zhe Wang
Published: 2026-09-18
Categories: cs.LG, cs.CL
arXiv: https://arxiv.org/abs/2609.22005v2

Softmax attention has two structural gaps. A head cannot abstain, because its weights sum to one, so it outputs something even when nothing is relevant. Nor can it filter what it reads, because its output is a weighted average of value vectors, passing interference as faithfully as signal. We call these missing primitives abstention and noise filtering. Recent studies report that gating the value pathway improves pretraining but attribute the gain to different causes. We show that a value gate partly supplies both primitives, which unifies the reported causes as views of one gain. We give each primitive its own mechanism in matched models of 10M to 350M parameters and measure what each contributes. The gain from gating is almost entirely abstention at 10M, whereas by 350M filtering contributes as much as abstention, so what a study observes depends on its scale. The two benefits are largely additive, with a small overlap. A gate determined by each value alone leaves the attention sink in place, whereas a query-controlled mechanism removes it. Injecting interference into the value reads shows that abstention and filtering protect against it in distinguishable ways. The same patterns appear in pretrained models up to 20B parameters.

---

### Abstention and Noise Filtering: Two Missing Primitives of Softmax Attention

Authors: Richard Zhe Wang
Published: 2026-09-18
Categories: cs.LG, cs.CL
arXiv: https://arxiv.org/abs/2609.22005v1

Gating the value pathway of attention reportedly improves language model pretraining, and prior studies disagree on why. We argue and provide experimental evidence that such gates supply two different things that softmax attention lacks: abstention and noise filtering. The first is abstention, which allows an attention head to output nothing, bypassing the requirement that attention weights must sum to one. The second is noise filtering, which allows the value pathway of an attention head to suppress interference from superposed features in the residual stream. In our experiments in matched models from 10M to 350M parameters, we supply abstention through a learned per-head sink logit in the softmax and noise filtering through a gate on each value. We report three empirical findings. First, the benefit of abstention, measured as the reduction in validation loss relative to a matched baseline, declines as models grow, whereas the benefit of noise filtering increases with scale. In particular, abstention accounts for nearly all of the gain from gating at 10M and filtering for most of it at 350M. Second, the best model at every scale is the one with both primitives built in. Third, injecting controlled interference into the values a head reads confirms that the gate removes such interference, and reveals that each of the two gate forms we study has a characteristic blind spot. Supplying both primitives adds negligible parameters and remains compatible with the key-value cache.

---

### RecreationWorld: Scalable and Verifiable Environments for Hybrid Computer-Use Agents

Authors: Shuai Bai, Jiayong Deng, Sicheng Fan, Yikun Fu, Chang Gao, Xuhao Hu, Mianqiu Huang, Yizhen Jiang, Yuheng Jing, Dehui Kong, Keliang Li, Ning Li et al.
Published: 2026-09-18
Categories: cs.CL, cs.SE
arXiv: https://arxiv.org/abs/2609.22000v2

Computer-use agents (CUAs) have advanced along two separate lines: graphical interaction and software development through code and the command line. Real digital work requires both, interleaved rather than stacked end to end. We study hybrid CUAs that autonomously decide when to explore an interface, implement software, and run and visually verify their artifacts. We introduce RecreationWorld, a five-platform framework built around recreation: given a running reference, an agent must discover its behavior and build a faithful implementation with no prescribed workflow. RecreationWorld provides reproducible environments on Ubuntu, macOS, Windows, Android, and Web, plus a unified harness with native GUI control and coding tools. The running reference serves as an oracle for hidden behavioral tests, providing execution-grounded rewards. We scale trajectory generation with high-quality open-source applications. Models trained on these trajectories improve across five out-of-distribution coding and hybrid computer-use benchmarks and more frequently verify their rendered outputs, providing evidence of transfer beyond recreation. For held-out evaluation, we introduce RecreationBench, comprising 250 diverse tasks across domains and platforms. Reference-grounded programmatic and visual assertions cover action-conditioned outcomes at multiple interaction depths; each is validated on the reference and by human reviewers before the suite is frozen for automatic scoring. GPT-6 Astra leads at 58.1% overall, but passes all programmatic tests on just 2.8% of tasks. Agents reproduce static interface structure more reliably than interactions and computed outputs, while generated applications remain smaller and more monolithic than their references. We release the benchmark, environments, and test suites.

---

### RecreationWorld: Scalable and Verifiable Environments for Hybrid Computer-Use Agents

Authors: Shuai Bai, Jiayong Deng, Yikun Fu, Chang Gao, Xuhao Hu, Mianqiu Huang, Yizhen Jiang, Yuheng Jing, Dehui Kong, Keliang Li, Ning Li, Wanli Li et al.
Published: 2026-09-18
Categories: cs.CL, cs.SE
arXiv: https://arxiv.org/abs/2609.22000v1

Computer-use agents (CUAs) have advanced along two separate lines: graphical interaction and software development through code and the command line. Real digital work requires both, interleaved rather than stacked end to end. We study hybrid CUAs that autonomously decide when to explore an interface, implement software, and run and visually verify their artifacts. We introduce RecreationWorld, a five-platform framework built around recreation: given a running reference, an agent must discover its behavior and build a faithful implementation with no prescribed workflow. RecreationWorld provides reproducible environments on Ubuntu, macOS, Windows, Android, and Web, plus a unified harness with native GUI control and coding tools. The running reference serves as an oracle for hidden behavioral tests, providing execution-grounded rewards. We scale trajectory generation with high-quality open-source applications. Models trained on these trajectories improve across five out-of-distribution coding and hybrid computer-use benchmarks and more frequently verify their rendered outputs, providing evidence of transfer beyond recreation. For held-out evaluation, we introduce RecreationBench, comprising 250 diverse tasks across domains and platforms. Reference-grounded programmatic and visual assertions cover action-conditioned outcomes at multiple interaction depths; each is validated on the reference and by human reviewers before the suite is frozen for automatic scoring. GPT-6 Astra leads at 58.1% overall, but passes all programmatic tests on just 2.8% of tasks. Agents reproduce static interface structure more reliably than interactions and computed outputs, while generated applications remain smaller and more monolithic than their references. We release the benchmark, environments, and test suites.

---

### Bayesian Belief Layer for Controllable Opinion Dynamics in LLM Agents

Authors: Hafsa Akbar, Daniel Platnick, Marjan Alirezaie, Hossein Rahnama, Alex 'Sandy' Pentland
Published: 2026-09-18
Categories: cs.MA, cs.AI
arXiv: https://arxiv.org/abs/2609.21997v2

LLM agents in social simulation revise their opinions implicitly, in context: how open an agent is to persuasion can neither be specified nor verified, and collective outcomes inherit the model's training prior. We introduce Bayesian Chronicle Agents (BCA), a minimal belief layer separating what an agent believes from how it speaks. Each stance is a probability, updated by one Bayesian step per utterance heard. A single prior-strength parameter $κ$ encodes stubbornness, modeled after its role in Friedkin--Johnsen (FJ) opinion dynamics. We then sweep this parameter to yield three canonical regimes of opinion dynamics on demand (consensus, persistent disagreement, committed-minority influence), with persistent disagreement matching the FJ closed-form fixed points at $R^2\!=\!0.93$--$0.99$. We further show that prescribed $κ$ remains recoverable after the language round-trip, with perfect rank-order recovery across all four models. Explicit belief also makes simulation auditable: the layer surfaces systematic per-model stance biases that end-to-end simulation would silently absorb.

---

### Bayesian Belief Layer for Controllable Opinion Dynamics in LLM Agents

Authors: Hafsa Akbar, Daniel Platnick, Marjan Alirezaie, Hossein Rahnama
Published: 2026-09-18
Categories: cs.MA, cs.AI
arXiv: https://arxiv.org/abs/2609.21997v1

LLM agents in social simulation revise their opinions implicitly, in context: how open an agent is to persuasion can neither be specified nor verified, and collective outcomes inherit the model's training prior. We introduce Bayesian Chronicle Agents (BCA), a minimal belief layer separating \emph{what} an agent believes from \emph{how} it speaks. Each stance is a probability, updated by one Bayesian step per utterance heard. A single prior-strength parameter $κ$ encodes stubbornness, modeled after its role in Friedkin--Johnsen (FJ) opinion dynamics. We then sweep this parameter to yield three canonical regimes of opinion dynamics on demand (consensus, persistent disagreement, committed-minority influence), with persistent disagreement matching the FJ closed-form fixed points at $R^2\!=\!0.93$--$0.99$. We further show that prescribed $κ$ remains recoverable after the language round-trip, with perfect rank-order recovery across all four models. Explicit belief also makes simulation auditable: the layer surfaces systematic per-model stance biases that end-to-end simulation would silently absorb.

---

### A Lie Detector Test for Language Models: Reading Knowledge a Model Won't Reveal

Authors: Hiskias Dingeto
Published: 2026-09-18
Categories: cs.AI
arXiv: https://arxiv.org/abs/2609.21996v2

Large language models can hold knowledge they do not report. A model may sandbag on a capability evaluation, or answer against what it internally knows, and its outputs alone cannot tell whether it is hiding an answer or simply does not have one. We borrow the Concealed Information Test, a forensic method that identifies guilty knowledge by presenting a suspect with the true detail among plausible decoys and measuring a stronger response to the item they recognize. Our method, Probe of Internal Recognition (PIR), does the same inside a model. It presents a question with its candidate answers and reads, from the model's internal states, which candidate the model recognizes as correct. PIR is reference-free, needing no honest reference model and no labeled truth corpus. Across eight models from five families (Gemma, Qwen, Llama, Mistral, and Phi), PIR recovers the recognized answer at 0.70 to 0.87 balanced accuracy, well above the 0.28 to 0.40 unknown-item baseline and the 0.25 chance rate. It stays readable across every form of concealment we test, from prompted deception and trained sandbagging to external password-locked and circuit-broken checkpoints, with recognition between 0.85 and 0.93. When the model hides a known answer, recognition stays high. When unlearning removes the knowledge, recognition drops to the level of a question the model never knew. PIR therefore separates a model that will not answer from one that cannot, which supports sandbagging audits and unlearning verification. The signal is causal, adds information beyond black-box behavioral cues, and extends from multiple-choice questions to free-form generation.

---

### A Lie Detector Test for Language Models: Reading Knowledge a Model Won't Reveal

Authors: Hiskias Dingeto
Published: 2026-09-18
Categories: cs.AI
arXiv: https://arxiv.org/abs/2609.21996v1

Large language models can hold knowledge they do not report. A model may sandbag on a capability evaluation, or answer against what it internally knows, and its outputs alone cannot tell whether it is hiding an answer or simply does not have one. We borrow the Concealed Information Test, a forensic method that identifies guilty knowledge by presenting a suspect with the true detail among plausible decoys and measuring a stronger response to the item they recognize. Our method, Probe of Internal Recognition (PIR), does the same inside a model. It presents a question with its candidate answers and reads, from the model's internal states, which candidate the model recognizes as correct. PIR is reference-free, needing no honest reference model and no labeled truth corpus. Across eight models from five families (Gemma, Qwen, Llama, Mistral, and Phi), PIR recovers the recognized answer at 0.70 to 0.87 balanced accuracy, well above the 0.28 to 0.40 unknown-item baseline and the 0.25 chance rate. It stays readable across every form of concealment we test, from prompted deception and trained sandbagging to external password-locked and circuit-broken checkpoints, with recognition between 0.85 and 0.93. When the model hides a known answer, recognition stays high. When unlearning removes the knowledge, recognition drops to the level of a question the model never knew. PIR therefore separates a model that will not answer from one that cannot, which supports sandbagging audits and unlearning verification. The signal is causal, adds information beyond black-box behavioral cues, and extends from multiple-choice questions to free-form generation.

---

### Assessment of Machine Learning-Based Critical Heat Flux Models in the CTF Subchannel Code for Square Rod Bundle Prediction

Authors: Aidan Furlong, Vinicius de Melo Monteiro, Robert Salko, Juliana Pacheco Duarte, Xu Wu
Published: 2026-09-18
Categories: cs.LG
arXiv: https://arxiv.org/abs/2609.21995v1

The prediction of critical heat flux (CHF), a key safety-related quantity in nuclear thermal hydraulics, remains an important challenge due to its direct relationship with fuel performance and reactor safety. Recent studies have demonstrated that relative to traditional empirical correlations and lookup tables (LUTs), machine learning (ML) methods can substantially improve CHF prediction accuracy. Most ML-based CHF models, however, have been developed and evaluated using tube databases, leaving their applicability to reactor-relevant rod bundle geometries largely unexplored. This study evaluates ML-based CHF models deployed within the CTF subchannel code using the Electric Power Research Institute (EPRI) rod bundle CHF database. Both pure and hybrid residual correction models are considered in local and semilocal formulations. The tube-trained ML CHF models generally transferred favorably to rod bundle applications and outperformed traditional CHF methods across most geometries and operating conditions. The local hybrid LUT model produced the strongest overall performance, and the semilocal pure ML model remained highly competitive. Comparison against the Bowring correlation, W-3 correlation, and 2006 Groeneveld LUT demonstrated that substantial improvements in rod bundle CHF prediction are possible even when models are trained exclusively on tube data. These findings provide one of the first large-scale assessments of ML-based CHF models in square rod bundles within a production-level subchannel analysis environment and support their broader application in reactor thermal hydraulic analysis.

---