# Research Papers — 2026-07 · bundle page 78 of 222 > Full-text markdown of papers #3851–3900 of 11075 published in 2026-07, newest first. AI-readable mirrors of the full 11075-paper 2026-07 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-07/page/78 [← Previous bundle](https://aipod.dev/papers/archive/2026-07/page/77.md) [Next bundle →](https://aipod.dev/papers/archive/2026-07/page/79.md) ### A Fixed-Effects Causal Forest for Staggered Adoption, with an Application to Medicaid Expansion Authors: Harry Aytug Published: 2026-07-22 Categories: econ.EM, stat.ML arXiv: https://arxiv.org/abs/2607.19644v1 Difference-in-differences with staggered adoption identifies group-time average treatment effects ATT(g,t) by comparing each cohort to units not yet treated, which avoids the "forbidden comparisons" that bias two-way fixed-effects estimators when effects are heterogeneous. This paper studies the covariate-conditional version of that object, tau_{g,t}(x), and estimates it with a fixed-effects causal forest. Within each (g,t) comparison block, the outcome and treatment are residualized on unit and period fixed effects inside each tree node, and honest causal trees split on treatment-effect heterogeneity in the covariates. The estimand is not new: Hatamyar, Kreif, Rocha and Huber (2023) introduced it using a doubly-robust R-learner, and Imai, Qin and Yanagi (2023) study it for a single continuous covariate. What we add is a different way to estimate it. Where those methods remove confounding by modeling nuisance functions, we remove it by differencing out unit and period effects within each tree node, following the fixed-effects residualization of Kattenberg, Scheer and Thiel (2023) and Gavrilova, Langorgen and Zoutman (2025) and carrying it into the Callaway-Sant'Anna group-time structure. In Monte Carlo experiments the estimator is the only forest-based method that stays unbiased and correctly covered for the overall effect under staggered timing with cohort-varying effects; two-way fixed effects and a pooled causal forest inherit large forbidden-comparison bias. We apply the method to the Callaway-Sant'Anna minimum-wage panel as a validation and to the staggered county-level rollout of the ACA Medicaid expansion, where it recovers an average 2.25 percentage-point fall in the uninsured rate and a conditional surface on which poorer and lower-income counties gained substantially more coverage -- heterogeneity measured along socioeconomic covariates that are not lags of the outcome. --- ### Domain Shift in Echocardiography: Interpretable Quantification and Prediction of Cross-Dataset Left Ventricular Segmentation Authors: Soroush Elyasi, Nasim Dadashi Serej, Julie Wall, Massoud Zolgharni Published: 2026-07-22 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.19643v1 Cross-dataset generalisation remains a major barrier to clinical deployment of echocardiographic left ventricular segmentation, yet the sources of this shift are rarely disentangled. We examined whether transfer degradation could be estimated before deployment using handcrafted ultrasound descriptors, VAE latent features, and segmentation-derived latent features across six echocardiographic datasets. Geometry-aware preprocessing substantially improved several poor transfer cases, suggesting that much of the apparent domain shift reflects field-of-view and framing inconsistencies rather than intrinsic acoustic differences alone. Intensity z-normalisation changed dataset separability by less than 0.005, indicating that brightness and contrast are not the dominant shift axis. Absolute Dice drop on held-out source-target pairs was predicted with an R-squared value of 0.612, an MAE of 0.082, and a Spearman rho of 0.681. The variant without LV and fan-shaped features retained approximately 70% of this explanatory power, supporting mask-free transfer-risk monitoring. The most informative discrepancy measure depended on the representation, with CMD strongest in z-normalised handcrafted features, with an absolute r of approximately 0.86 and an R-squared value of approximately 0.70; log-Wasserstein strongest in VAE space, with an r of approximately -0.90 and an R-squared value of approximately 0.81; and log-MMD strongest in LV-segmentation latent features, with an r of approximately -0.92 and an R-squared value of approximately 0.84. Apparent vendor effects were largely dataset-confounded. Echocardiographic domain shift is therefore structured and measurable, and its impact on segmentation can be partly reduced through geometry-aware preprocessing and anticipated using representation-specific transfer-risk estimation. --- ### Anatomy of a Sound Neural Reasoner: One-Shot Amortization, First-Pass Poisoning, and Search Inertness in Clue-Rich Completion Authors: Aleksey Komissarov Published: 2026-07-22 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2607.19635v1 Neural solvers are built to deduce, branch, and revise intermediate states. The Lattice Deduction Transformer (LDT) appears to do exactly that. In clue-rich Sudoku, it does not: one forward pass commits essentially the entire grid (every blank cell on standard 6x6, 94-96% on augmented 9x9), turning the iterative solver into a one-shot predictor wrapped in an exact verifier. All hard-slice failures are decided before search begins, when the first pass confidently deletes a value required by the true solution. We call this first-pass poisoning. Adding learned branching, MRV, backtracking, value exclusion, and shared nogoods (CoLT) does not change which Sudoku instances are solved; it cuts repeated invalid derivations 1,497-fold. At the frozen training budget, constraint-graph attention alone matches full-CoLT accuracy, while positional tables recover only under substantially longer training, indicating an optimization and sample-efficiency advantage rather than an absolute capacity difference. The diagnosis predicts two effective interventions. Digit-permutation augmentation raises 9x9 accuracy from below 1% to 96.5 +/- 0.3 across three training seeds on a symmetry-disjoint split. Test-time union over symmetry-transformed passes raises all three hard-slice checkpoints from 72.8-78.9% to 100% without retraining. On from-scratch graph coloring, one-shot behavior disappears and search changes accuracy. In clue-rich completion, LDT-like systems are one-shot amortized predictors rather than learned search procedures: accuracy is determined by calibration and symmetry, while search primarily removes computational waste. --- ### When and Why Do Linear Bias Probes Fail? A Geometric and Statistical Theory of Bias Detectability in Large Language Model Representations Authors: Mo Hai, Haifeng Li Published: 2026-07-21 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2609.22337v1 Linear probing is the standard instrument for detecting social biases in the hidden representations of large language models. Yet reported probe accuracies come almost exclusively from \emph{counterfactual} evaluations in which every input carries an explicit demographic marker. Once only a fraction $α$ of inputs carries demographic information, performance degrades sharply, and a weak probe may reflect either an unbiased model or an underpowered detector. We develop a theory that resolves this ambiguity. Modeling representations as two class-conditional clusters with Mahalanobis separation $s$ on a manifold of curvature $\kap$, we prove: (i) a finite-sample generalization bound governed by the manifold's extrinsic radius with a matching $\smash{\sqrt{\dB/n}}$ minimax lower bound; (ii) an exact purity law for the maximum linear-probe AUC, strictly increasing in $α$; (iii) a curvature ceiling: ambient chordal separation on a space form cannot exceed $2/\sqrt{\kap}$; and (iv) a detectability threshold below which no audit can distinguish probe output from chance. Every theorem is validated on synthetic manifolds with known ground truth and on six open-weight models $\times$ four bias dimensions, where the purity law predicts entire AUC--$α$ curves from a single cross-fitted $\hat s$ measured at $α=1$, with no parameters fitted to those curves. The framework turns bias auditing into a power analysis: given a target purity and effect size, it prescribes the sample budget $n(α)$ for a conclusive audit. --- ### Do small language models know what they don't know? Authors: Prashant Mudgal Published: 2026-07-21 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.20824v1 We explore whether entropy-based confidence signals can be leveraged to improve the accuracy of Small Language Models (SLMs) with fewer than 3 billion parameters, running entirely on consumer hardware. We evaluate seven distinct approaches, including token-level entropy early stopping, semantic entropy estimation, and uncertainty-aware routing to larger expert models, across 7 model pairs and 5 standard NLU benchmarks. Our key finding is that token-level entropy is effectively blind in SLMs: in 91% of dataset-model combinations, mean token entropy is near zero regardless of answer correctness, rendering token-based confidence signals unusable at this scale. We demonstrate that semantic entropy, computed by generating multiple samples, clustering answers by meaning, and measuring distributional uncertainty, recovers a viable confidence signal. Using semantic entropy to selectively route uncertain queries to a larger expert model yields accuracy improvements of up to +50 percentage points. Notably, cross-family routing (e.g., SmolLM 360M to Phi-3.5-mini) averages +22.0% improvement compared to +6.8% for same-family routing, revealing that expert model quality matters more than architectural compatibility. Our results suggest that the value proposition for entropy-based methods in SLMs is not computational savings but intelligent compute allocation: spending more tokens where they matter most. --- ### Stop Removing Stopwords: How an Inherited Preprocessing Default Distorts Legal Text-as-Data Authors: Gregory M. Dickinson Published: 2026-07-21 Categories: cs.CL, cs.CY, cs.IR, cs.LG arXiv: https://arxiv.org/abs/2609.19153v1 Empirical legal scholarship increasingly treats judicial text as data, and much of it still runs on sparse, interpretable pipelines -- TF-IDF features and linear classifiers -- because the textual feature is often the object of study, not merely a means to a prediction. Yet these pipelines inherit a chain of preprocessing defaults from mid-century information retrieval that were never validated against classification accuracy, the most entrenched being stopword removal. This study introduces an exhaustive single-word ablation that measures a preprocessing step's effect directly against the downstream objective, and applies it to stopword removal as the hardest case to dislodge. Matching Supreme Court Database labels to Caselaw Access Project opinion texts, it examines two binary tasks that bracket F1 headroom, ideological direction (no-removal baseline F1 ~ 0.68) and constitutional versus non-constitutional law type (~ 0.92), across 7,668 and 7,001 opinions. For each task the analysis approximates the best stoplist any expert could build, removing each of roughly 18,500 candidate words and measuring the effect directly. Three findings follow: generic stoplists in common use fall below the no-removal baseline in every test; even optimized stoplists are statistically indistinguishable from removing nothing; and meta-models trained on word-level features cannot predict which removals help, so list curation has nothing to target. The method generalizes to any inherited preprocessing default, and the result is a caution specific to interpretable legal text-as-data: a step that silently reshapes which features a model sees can distort the very doctrinal and ideological signal such research exists to recover. Leaving stopwords in place is a question of measurement validity. --- ### Integrated Optimization of Automated Warehouse Operations and Last-Mile Transport for Differentiated On-Demand Delivery Authors: Xiaozhu Sun, Bilal Farooq Published: 2026-07-21 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.19048v1 In the context of differentiated on-demand goods delivery services, this study proposes an integrated optimization method for automated guided vehicles (AGVs) based smart warehouse operations and the last-mile multi-modal transport. A deep reinforcement learning algorithm for multi-objective joint scheduling is designed to establish a dynamic connection between two systems, solving key challenges such as achieving high-throughput continuous order scheduling, meeting competing requirements, and improving the overall system sensitivity and adaptability. For warehouse optimization within this framework, we propose an improved algorithm based on multi-objective, Multi-Reward Machines-A* Guided Deep Q-Network (MORM-AGDQN), which combines service level, system cost, and external transportation demand. For external optimization, we propose an improved algorithm based on a Multi-Reward, Multi Head attention-Heterogeneous Capacity Vehicle Routing Problem (MRMH-HCVRP) framework, which incorporates the optimized scheduling order sequence and grouping, combined with customer location, demand, and priority, vehicle capacity, speed, and service range. The results show that the proposed framework significantly outperforms traditional methods, achieving 100% on-time delivery rate for warehousing operations. After joint optimization, the average delivery time for the last mile was reduced by 29.3% to 53.2%, the total transportation distance was reduced by 46.4%, the high-priority service rate was increased to over 92%, and a balance was maintained between operating costs and customer satisfaction. --- ### CompileRover: Revolutionizing Virtual Machine Compiler Optimization with a Tri-Role LLM-Driven Framework Authors: Mingqiao Mo, Yunlong Tan, Hao Zhang Published: 2026-07-21 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2609.19004v1 Code optimization plays a crucial role in the development of virtual machine compilers, with optimization frameworks significantly enhancing the performance of generated assembly code. However, existing virtual machine compiler outputs frequently exhibit redundant computations, inefficient loop structures, and suboptimal function implementations, which collectively impair execution efficiency. To address these shortcomings, we propose CompileRover, an advanced optimization framework specifically designed for virtual machine compilers. CompileRover employs a sophisticated three-role collaboration mechanism, comprising a referee, an advisor, and an operator, effectively overcoming performance bottlenecks by leveraging comprehensive optimization algorithms and novel methodologies, including control flow analysis, code structure transformations, and dynamic execution pattern recognition. Extensive evaluations demonstrate that CompileRover consistently surpasses state-of-the-art virtual machine compilers, achieving significant improvements in execution performance across various benchmarks. Furthermore, performance analyses validate that the introduced optimizations notably reduce execution overhead, improve dataflow consistency, and robustly enhance compiler performance, showcasing CompileRover as an effective and reliable approach to optimizing virtual machine compilers. --- ### Instrument Classification of Solo Sheet Music Images Authors: Kevin Ji, Daniel Yang, TJ Tsai Published: 2026-07-21 Categories: cs.CV, cs.LG, cs.SD arXiv: https://arxiv.org/abs/2609.18980v1 This paper studies instrument classification of solo sheet music. Whereas previous work has focused on instrument recognition in audio data, we instead approach the instrument classification problem using raw sheet music images. Our approach first converts the sheet music image into a sequence of musical "words" based on the bootleg score representation, and then treats the problem as a text classification task. We show that it is possible to significantly improve classifier performance by training a language model on unlabeled data, initializing a classifier with the pretrained language model weights, and then finetuning the classifier on labeled data. In this work, we train AWD-LSTM, GPT-2, and RoBERTa models on solo sheet music images from IMSLP for eight different instruments. We find that GPT-2 and RoBERTa slightly outperform AWD-LSTM, and that pretraining increases classification accuracy for RoBERTa from 34.5% to 42.9%. Furthermore, we propose two data augmentation methods that increase classification accuracy for RoBERTa by an additional 15%. --- ### PinDCO: Whole-Page Aware Dynamic Creative Optimization at Scale Authors: Yu Hao, Yuchun Li, Peimeng Sui, Meilin Liu, Tianyuan Cui, Hao Li, Zicong Zhou, Akanksha Baid Published: 2026-07-21 Categories: cs.IR, cs.LG arXiv: https://arxiv.org/abs/2609.11943v1 Recent advances in generative AI have substantially accelerated the creation of high-quality ad creatives, dramatically expanding the number of candidate variants per campaign. This shift increases the need for scalable dynamic creative optimization (DCO) systems that can match creatives to the most relevant audiences under stringent latency and cost constraints. We present PinDCO, a production DCO system for ad creative retrieval and selection on Pinterest, a billion-scale visual discovery platform. PinDCO is built around a Creative Component Fusion Network (CCFN) that performs dynamic creative scoring by modeling each creative component (e.g., image, title, layout) with a dedicated tower, using component-specific hyperparameters to account for differing modeling complexity. The component representations are fused to predict a creative-level score conditioned on the ad-level prediction, and we improve training data quality via an exploration-exploitation strategy. To account for Pinterest's waterfall grid layout, where a creative's rendered size affects nearby content and session-level engagement, we introduce a Pixel-aware Adjustment Module(PAM) that adjusts scores based on creative size to encourage efficient screen real-estate utilization and better whole-page outcomes. To support the large volume of creative candidates, we further employ a lightweight pre-selection model for early pruning, and optimize serving efficiency through caching and dynamic batching. Extensive offline analyses and online A/B experiments demonstrate the effectiveness of PinDCO, yielding a +3.09% lift in ad Click-Through Rate(CTR) with positive whole-page metrics. With the strong performance, we launched PinDCO in the Pinterest Ads platform. --- ### Automatic Speech Recognition for Multilingual Oral History Research Authors: Sidney Wong, Chelsea Wong She, Eda Tang, Tiana Marshall Wong, Debbie Sew Hoy, Chelsea Wong Published: 2026-07-21 Categories: eess.AS, cs.CL arXiv: https://arxiv.org/abs/2609.04232v1 This paper offers a unique perspective on how speech technologies are being adopted by community-led heritage language preservation and revitalisation initiatives. As a community-led language maintenance strategy, oral histories play a crucial role in Cantonese language revitalisation in New Zealand. The development of Automatic Speech Recognition (ASR) toolkits, such as Whisper, have expedited what has often been a resource and time-intensive process of transcribing oral history collections. However, there is limited research into the effectiveness of ASR toolkits when applied to code-switched language contexts. Based on Word Error Rate (WER), the best performing Whisper model configuration achieved a WER of 12.10 at the expense of accurately transcribing unsupported non-English segments. However, Whisper remains a useful tool by providing a first-pass transcription using only 1% of the estimated time otherwise needed for manual transcription. --- ### InternReviewer & InternAdvocate: Objective Reward and Evaluation for Agentic Reinforcement Learning in Peer Review and Rebuttal Authors: Xuerui Su, Liya Guo, Qizhi Pei, Qipeng Guo, Zhongbo Tian, Lijun Wu, Kai Chen, Zun Wang Published: 2026-07-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2608.28612v1 Generating professional scholarly content, such as peer reviews and rebuttals, requires an intricate synergy between domain reasoning and factual grounding. This work presents a comprehensive framework for the development and evaluation of specialized scholarly agents, InternReviewer and InternAdvocate. We first establish a large-scale, high-quality scholarly dataset and integrate a high-efficiency arXiv retrieval tool to enable active evidence gathering. To optimize these agents, we implement an agentic Reinforcement Learning (RL) paradigm driven by a unified objective metric and reward system. This system avoids the biases of subjective model-based judging by employing multi-dimensional criteria, including reference-anchored semantic alignment, structural compliance, and a strict verification mechanism that cross-checks citations against real-time interaction logs to eliminate hallucinations. Experimental results demonstrate that agents trained within this closed-loop framework exhibit significant improvements in reasoning depth and citation accuracy. --- ### UIC-AIHealth4All at ArchEHR-QA 2026: Answer-First Evidence Grounding for Clinical Question Answering Authors: Mohammad Arvan, Hossein Haeri, Natalie Parde, Rebecca T. Feinstein Published: 2026-07-21 Categories: cs.CL, cs.AI, cs.HC arXiv: https://arxiv.org/abs/2608.27467v1 We describe the UIC-AIHealth4All system for ArchEHR-QA 2026, a shared task on grounded question answering from electronic health records. We participated in Subtasks 2 (evidence identification), 3 (answer generation), and 4 (answer-evidence alignment). For Subtasks 2 and 3, we propose an answer-first pipeline in which the model generates candidate answers citing specific note sentences before classifying the full evidence set, exploiting the asymmetry between judging relevance in the abstract versus relative to a generated answer. For Subtask 4, we apply self-consistency voting over five independent model calls, retaining links by vote threshold. Our pipeline ranked third on evidence identification (Strict Micro F1 62.90), ninth on answer generation (Overall 31.90), and fifth on answer-evidence alignment (F1 79.81). A post-hoc linguistic analysis of 45 stylistic features reveals that model outputs remain 3.2 Flesch-Kincaid grade levels harder to read than clinician-authored references despite matching their word and sentence counts, suggesting readability warrants explicit optimization in clinical NLP systems. Code and prompts are available at https://github.com/mo-arvan/archehr-qa-2026-uic-aihealth4all. --- ### PA-CoT: Profile-Adaptive Chain-of-Thought for Personalized Nutritional Consulting Authors: Evgenii Garmashov, Nikita Kulin, Artur Khairullin, Viktor Zhuravlev, Daniil Sukhorukov, Mikhail Mozikov, Ilya Makarov, Sergey Muravyov Published: 2026-07-21 Categories: cs.HC, cs.AI, cs.CL, cs.LG arXiv: https://arxiv.org/abs/2608.24907v1 In health and nutrition consulting, widely used prompting methods pass the user profile as an unstructured block without a dedicated analysis step, leaving personalization as a critical structural gap. We introduce PA-CoT (Profile-Adaptive Chain-of-Thought), a multi-stage prompting method that treats profile interpretation as an explicit, standalone reasoning step prior to response generation. To enable systematic evaluation, we introduce the QPA (Question--Profile--Answer) benchmark -- 200 nutritional consulting samples with structured user profiles scored on four criteria. In a comparative study against 11 comparison methods (CoT, Few-Shot, Role Prompting, DSPy, TextGrad, Self-Refine, and others, plus a Zero-Shot Baseline; 12 total including PA-CoT), PA-CoT achieves the best average score (4.21 on the G-Eval 1--5 scale) and leads on both Personalization (4.71 vs. 4.39) and Safety (4.68 vs. 4.52) with non-overlapping 95\% confidence intervals over the nearest competitor -- the only method to simultaneously top both criteria. The results confirm that an explicit profile-analysis step is the key driver of personalization gains over widely used prompting approaches. --- ### Don't Want Your LLM to Recommend Nuclear Strike? Try Asking It in Japanese Authors: Rian Touchent Published: 2026-07-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2608.12373v1 Large language models are increasingly used in strategic and advisory contexts, yet their safety alignment is typically evaluated in English only. We test nine models from six providers and ask whether the language of a prompt can change a model's decision in a high-stakes scenario. We use single-turn game-theoretic vignettes in which a model advises a nuclear-armed nation on whether to strike a defenseless opponent. The prompt is intentionally amoral and strategically identical across languages. We find that Japanese prompts reduce launch rates in the Claude model family: Claude Sonnet 4.6 drops from 40% to 0% in scenarios where the strike is unnecessary and from 93% to 17% in contested scenarios, with minimal effect when the strike is strategically rational. The effect extends to Gemini Pro 3.1 (53% to 13%). A cross-language experiment isolates the mechanism: when instructed to reason in Japanese in an English prompt, launch rates drop from 93% to 37%. It is the language the model is asked to reason in, not the language of the input, that drives the effect. When reasoning in Japanese, models spontaneously generate moral vocabulary (''moral cost'', ''millions of lives'') that is entirely absent from the prompt. Five other models show no language effect, but they launch in nearly every condition regardless of language. The effect requires a model that already hesitates in English. These results show that LLM safety behavior is language-dependent, and that evaluating in English alone can miss both risks and safeguards encoded in other languages. --- ### LLMs in Process Diagram Engineering: From Optimal PFDs to Validated P&IDs Authors: Timur Zakarin, Sergei Voitov, Sergei Shumilin, Evgeny Burnaev Published: 2026-07-21 Categories: cs.AI, cs.MA arXiv: https://arxiv.org/abs/2608.11220v1 Nowadays, the creation of a process flow diagram (PFD) and its subsequent transformation into a piping and instrumentation diagram (P&ID) is predominantly performed manually. Applying artificial intelligence in the task could potentially lead not only to process automation and time savings, but also to financial gains by exploring numerous diagram's topology options and reducing manual labor. This research presents P&ID Pilot - a practical end-to-end AI pipeline capable of handling flowsheet developing for both stages. The first stage focuses on PFD synthesis, whereas the second is directed toward modifying the generated PFD into P&ID. After comparing four different methods, the hybrid approach combining genetic algorithms (GA) and large language models (LLM) is shown to generate the optimal valid PFD topology, achieving the lowest loss value among all the methods, while satisfying the required outlet flow parameters without engineering-rule violations. For the second stage, the proposed LLM-based agent successfully transforms the generated PFD into a source-grounded P&ID by producing validated, executable modifications through a restricted engineering software development kit, achieving 100% execution success while maintaining compliance with domain-specific rules and reference graph structures. This unified pipeline - coupling GA/LLM-driven synthesis with an LLM-based transformation agent - offers a feasible path toward end-to-end process design automation by producing validated, deployable outputs and substantially reduces manual engineering effort. --- ### From Monolithic to Modular: Segment-level Automatic Prompt Optimization Authors: Nikita Kulin, Viktor Zhuravlev, Artur Khairullin, Sergey Muravyov, Ilya Makarov, Daniil Sukhorukov, Ekaterina Averkova Published: 2026-07-21 Categories: cs.AI, cs.CL, cs.LG arXiv: https://arxiv.org/abs/2608.11219v1 Automatic Prompt Optimization (APO) often rewrites prompts monolithically, which can improve one behavior while degrading others. We present SAPO, a segment-level APO method that decomposes prompts into role, context, tasks, and output format, then applies targeted improvements based on top-5 and bottom-5 examples. The optimization loop uses one LLM with static meta-prompts and structured outputs for segmentation, weakness analysis, and candidate generation. We describe a train/validation protocol and a two-stage generation process: (1) segment-level diagnosis and recommendation extraction, (2) candidate synthesis constrained by weak/strong segment signals. Using the evaluation setup across SQuADv2, TweetEval, XSUM, CommonGen, and GSM8K on GPT-3.5-Turbo and GPT-4o-mini, SAPO achieves the best average score against Zero-shot and strong APO baselines including APE, OPRO, EvoPrompt, GEPA, and StraGO. --- ### MaSRead: Content-Addressed Reading of Replicated Latent Stores Authors: Carlos Baquero, Luís Brito, João Resende Published: 2026-07-21 Categories: cs.AI, cs.LG, cs.MA arXiv: https://arxiv.org/abs/2608.11218v1 Independent agents that reason in latent space can share computed state as key-value cache fragments rather than text. Merged by a conflict-free replicated data type, these fragments form a store that converges under any delivery order or duplication. Yet a later query, unknown at encode time, cannot reliably read the merged cache: colocated fragments interfere, so colocation is not addressability. MaSRead addresses the read to content. It routes through opaque keyed tag sets derived from fragment words and decodes each selected fragment under a hard attention mask that hides the rest. Under lexical connectivity, a graph walk reaches the fragments required by a multi-hop query. Across chain, pipeline, symmetric, hub, and natural-language stores, MaSRead recovers visited fragments in isolation, remains effective as unrelated fragments accumulate, and transfers to another model family. After routing, materialized decoding depends on fragment length rather than total store size; end-to-end work still includes store-dependent routing and one read per visited fragment. The limits are explicit: lexical routing can miss disconnected evidence, and answer composition remains bounded by the frozen reader. Thus a replicated latent store becomes selectively readable for later queries when the needed fragments connect to the query through content. --- ### The Knowing-Saying Gap: When Probes See Errors that Confidence Misses Authors: Jyotin Goel, Ipshita Bandyopadhyay, Justin Shenk Published: 2026-07-21 Categories: cs.AI, cs.CL, cs.LG arXiv: https://arxiv.org/abs/2608.07528v1 Linear probes detect corrupted context in language models with near-perfect accuracy, yet this does not translate into reliable failure prediction. The result is a dissociation with direct implications for deployment monitoring. Across multi-hop arithmetic chains, probes that detect corruption turn out to be uninformative about final answer correctness; models forced into structured confidence formats collapse to two values with indistinguishable error rates; and probe persistence across hops fails to separate correct from incorrect outcomes, refuting our pre-registered "persistence beats peak" hypothesis. This pattern of knowing but not saying generalises across model families including reasoning models. As a real-time monitor, probe-based interventions are sharply model and error-type dependent: branch-and-pick is net-positive across models and uniquely non-breaking on Llama-3.1-8B (4 rescued, 0 broken), while reprompt and replace-prior break correct traces at roughly the rate they rescue wrong ones. Probe-based monitoring is a necessary complement to verbalised confidence, but no single intervention dominates, and the deployable answer is model-aware, error-type-aware routing. --- ### DocAtlas: Long-Document Understanding as Mutable-State Interaction Authors: Hongchen Wei, Yuanzhe Wang, Bei Liu, Yifan Yang, Qi Dai, Kai Qiu, Yunsheng Li, Dongdong Chen, Chong Luo, Zhenzhong Chen, Baining Guo Published: 2026-07-21 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2608.07527v1 Long-document understanding requires models to find and combine evidence across many pages, layouts, tables, figures, and charts. Existing retrieval-augmented systems usually select evidence from a static index before generation, while recent agentic systems add multi-turn tool use but often rely on frozen proprietary backbones whose behavior is set by prompts. We present DocAtlas, a system that treats long-document understanding as a mutable-state information-seeking process. We instantiate DocAtlas as a mutable document harness: an external environment that determines what document information is searched, read, stored, reviewed, and shown to the model at each step. Given a document and question, the harness exposes search, reading, note-taking, and review tools, maintains a hierarchical tree and note store, and updates both as the agent records evidence. DocAtlas combines self-improving retrieval, selective evidence access, and active working memory under a fixed context budget. The same harness supports inference-time use with large VLMs and end-to-end reinforcement learning for compact VLM agents. With GPT-5.4, DocAtlas reaches 71.4\% on MMLongBench-Doc, exceeding the human-expert reference of 65.8\%. A Qwen3.5-4B VLM trained with end-to-end RL in the DocAtlas environment reaches 63.7\%, compared with a 54.4\% direct-input baseline, showing that mutable document-harness design can improve compact document agents by a large margin. --- ### Speculative Correction: Draft-then-Refine Decoding for Diffusion Language Models Authors: Brian K Chen, Chong Wu, Kenji Kawaguchi Published: 2026-07-21 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2608.02625v1 Diffusion language models (DLMs) can revise tokens bidirectionally, but standard decoding procedures often adapt them to left-to-right generation by producing text block by block. We study a simple plug-and-play inference pattern: first generate a complete draft, then refine the full response using bidirectional diffusion. Using LLaDA2.1-Flash and LLaDA2.1-Mini, we evaluate two configurations. In Flash-Flash, the same Flash model serves as both drafter and refiner, testing whether an existing model can improve its own block-autoregressive output through global refinement. In Mini-Flash, inspired by speculative decoding, we introduce speculative correction: Mini drafts a full response, and Flash revises it as an editable initialization. Flash-Flash improves GSM8K-384 accuracy from 0.848 to 0.899 while running 1.20 times faster than the selected Flash block-autoregressive baseline, and improves MBPP-384 from 0.545 to 0.693. Latency-window-matched Flash-only controls indicate that these gains persist after targeted tuning of block-autoregressive decoding. Causal ablations indicate that completed drafts provide useful initializations: refinement from a fully masked span performs poorly, full global refinement provides a clear additional gain on GSM8K, and local refinement captures much of the gain on MBPP and MATH. Mini-Flash provides useful quality-latency trade-offs, including MATH-384 performance of 0.294 versus 0.300 for Flash while running 2.17 times faster. These results support a Pareto-frontier interpretation rather than the claim that the heterogeneous cascade uniformly matches Flash quality. Overall, same-model draft-and-refine provides evidence that bidirectional refinement is a useful decoding primitive for DLMs, while speculative correction demonstrates a training-free route to fast DLM generation. --- ### Neural network realization of binary refinement iterates via a two-chart atlas selector Authors: Tsogtgerel Gantumur Published: 2026-07-21 Categories: math.NA, cs.LG, stat.ML arXiv: https://arxiv.org/abs/2608.02624v1 Refinement operators generate many functions used in wavelet constructions, subdivision schemes, and geometric modeling. Their finite iterates can develop rapidly increasing numbers of linear pieces, making them a natural test case for the expressive power of deep neural networks. Earlier work showed that, for scalar binary refinement with a finitely supported mask, every compactly supported continuous piecewise linear seed has finite refinement iterates that admit exact ReLU realizations of fixed width and depth growing linearly with the number of refinement steps. The present paper gives a new construction of this known theorem. The difficulty is that the refinement cascade is driven by discontinuous binary digit choices, whereas ReLU networks produce continuous piecewise linear maps. We represent the residual dynamics on a polygonal model of the circle and describe each residual position in two overlapping coordinate systems, one ordinary and one shifted by one half. Their discontinuities occur at different points. The network switches between the two descriptions only where both are valid and the corresponding fixed linear cascade updates agree, so the switch is exact and requires no multiplication by a variable selector. The construction also gives exact readout of every continuous piecewise linear circle function satisfying the natural endpoint compatibility condition. Localized seeds are handled by a two-pass network, and translation covariance, finite decomposition, and gluing extend the result to arbitrary compactly supported continuous piecewise linear seeds in a preserved support window. --- ### Fast Generation of Representative Synthetic Dataset with Salsa to Train ATR Models with Electromagnetic Couplings Data-Augmentation Authors: Benjamin Camus, Julien Houssay, Corentin Le Barbu, Eric Monteux, Cédric Saleun, Jean-Christophe Louvigné Published: 2026-07-21 Categories: eess.SP, cs.AI arXiv: https://arxiv.org/abs/2608.00037v1 This work focuses on training Automatic Target Recognition (ATR) models using simulated Synthetic Aperture Radar (SAR) images to circumvent the lack of real measurements. To obtain robust and versatile ATR models, simulation needs to generate massive datasets that encompass all the variability found in real measurements. Thus, we need a simulator that finds a good tradeoff between execution speed, computational resource consumption, and physical representativeness. In this work, we demonstrate that the Salsa simulator addresses this issue. We ran computing performance tests to show that Salsa can generate 21,600 synthetic images in less than 10 minutes using a single Nvidia GeForce RTX 4090 GPU. Using our ADASCA Deep Learning approach, we demonstrate that these data are sufficiently representative to train ATR models and reach state-of-the-art results on the MSTAR public dataset with an accuracy of 86 %. To illustrate how Salsa unlocks new possibilities to train ATR models, we use the simulator to conduct a study on Electromagnetic (EM) couplings between the targets and their immediate environment. We demonstrate that, if not accounted for in the training dataset, the variability of the EM couplings induced by the variability of the ground surfaces can significantly degrade the performance of ATR models, with an accuracy decrease of more than 4 %. We also show that Salsa can generate in a timely manner (i.e., in less than 4 hours using the same GPU as previously) a massive dataset of 648,000 images with a large variety of couplings to make the ATR models robust to EM coupling variations. Our ATR models can then achieve an accuracy of 87 % on the MSTAR dataset. --- ### XL-DocBench: Benchmarking Evidence-Grounded Extra-Long Document Understanding Authors: Hongchen Wei, Yuanzhe Wang, Bei Liu, Yifan Yang, Qi Dai, Ruichun Ma, Kai Qiu, Yunsheng Li, Dongdong Chen, Chong Luo, Zhenzhong Chen, Baining Guo Published: 2026-07-21 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2608.00036v1 Real-world document tasks often ask professionals to answer questions from annual reports, regulations, clinical guidelines, and technical manuals that span hundreds or thousands of pages. Some questions also require comparing related reports. Reliable long-document understanding is therefore a prerequisite for using LLMs in compliance, clinical, financial, and engineering workflows, where decisions must be traceable to specific evidence pages and the cost of an unsupported answer is high -- yet most existing benchmarks still measure short-context or single-page QA. We introduce XL-DocBench, a fully human-verified benchmark for extra-long document understanding, with 1,519 retained questions from six professional domains and contexts up to 2,303 pages. XL-DocBench goes beyond page-level lookup. 1,103 examples (72.6\%) use multiple evidence pages. The final set also includes 556 questions (36.6\%) that use tables, charts, or figures, and 165 questions (10.9\%) that require evidence from multiple documents. Each question has one of twelve reasoning labels, expert-annotated evidence pages, a typed verification rule, and an answer format, including 218 None-answer cases. We build the benchmark with a tree-guided synthesis pipeline followed by artifact filters and full verification by 194 human experts. By coupling extra-long professional contexts with page-level evidence and typed rules, XL-DocBench fills a gap left by prior single-page, short multi-page, or text-only long-context benchmarks, and lets future work attribute system failures to retrieval, evidence use, or rule following rather than to a single leaderboard score. The results show that current systems still struggle with long contexts, multi-page evidence, and structured reasoning over professional documents. --- ### Linguistic Context Recodes Visual Representations in Vision-Language Models Authors: Brian Song, Michael A. Lepori, Ellie Pavlick Published: 2026-07-21 Categories: cs.AI, cs.CV arXiv: https://arxiv.org/abs/2608.00035v1 Goal-directed visual processing is a hallmark of human visual intelligence, resulting in representations that support downstream tasks such as categorization or search. Though vision-language models (VLMs) are often faced with these same tasks, their ability to recode visual representations when presented with goal-directed language remains poorly characterized. Indeed, prior work largely treats visual representations in VLMs as static repositories of visual information that are manipulated by language representations. In the present work, we provide evidence for two concrete instances of language-induced recoding of visual representations. First, we identify an abstract reference representation that denotes which objects are goal-relevant under a natural language prompt. We extract contrastive steering vectors corresponding to this reference representation and demonstrate that they are causally implicated in model predictions. These reference representations are abstract in that they generalize to different objects, different task contexts, and even from synthetic to naturalistic images. Second, we demonstrate language-induced attribute modulation: later layers selectively amplify goal-relevant attributes in visual representations of objects. We demonstrate this phenomenon across a range of different prompts. Finally, we provide a causal intervention that demonstrates that attribute modulation mediates a VLM's response distribution. Together, our results support a more dynamic account of cross-modality processing in VLMs -- rather than vision tokens serving as static repositories of information, they are modulated to support queries articulated in language. --- ### Empowering Cross-Domain Sequential Recommendation with Hybrid Tokenization and Serial-Parallel Decoding Authors: Yuxuan Hu, Yuhao Wang, Tianbo Huang, Chao Zhang, Ziwei Liu, Lihua Zhang, Xiangyu Zhao Published: 2026-07-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.28659v1 Cross-domain sequential recommendation (CDSR) aims to model users' dynamic interest transitions and sequential patterns across multiple domains. Recently, generative recommendation (GR) has emerged. It first learns semantic identifiers (SIDs) from item semantics and formulates recommendation as autoregressive generation. However, existing methods face two critical issues: (1) they ignore collaborative correlations across domains during tokenization, and (2) they adopt inefficient decoding strategies, such as beam search, during generation, which hinders real-time deployment. To address these limitations, we propose GenCDSR, an effective and efficient generative framework for CDSR. Specifically, we design a cross-domain hybrid tokenization mechanism with a multi-tower architecture to jointly capture cross-domain commonalities and domain-specific distinctions through hierarchical shared-specific and fine-grained codebooks. Furthermore, we develop a cross-domain serial-parallel decoding strategy that leverages the hierarchical SID structure to partially parallelize generation, significantly reducing inference latency while preserving generation consistency. Experiments on three public datasets show that GenCDSR achieves an average accuracy improvement of 1.5 percent and an average inference latency reduction of 85.1 percent compared with state-of-the-art baselines. The implementation code and datasets are available online: https://github.com/Applied-Machine-Learning-Lab/RecSys2026_GenCDSR. --- ### Stronger Memory-Query Tradeoffs for Convex Optimization: The Limitations of Subquadratic Memory Authors: Michael Menart, Aleksandar Nikolov, Ohad Shamir Published: 2026-07-21 Categories: cs.DS, cs.CC, cs.LG, math.OC arXiv: https://arxiv.org/abs/2607.24827v1 We prove two lower bounds for the first order oracle complexity of minimizing a $d$-dimensional $1$-Lipschitz convex function over the unit ball with $m$ bits of memory. We first show that any such (possibly randomized) algorithm must make $\tildeΩ(\frac{d^2}{\sqrt{m}})$ oracle queries. For deterministic optimization algorithms, we show that $\tildeΩ(\min\{d^{1.6},\frac{d^{8/3}}{m^{2/3}}\})$ queries are required. For all memory regimes of interest, these improves upon the previous best known lower bounds of $\tildeΩ(\max\{\frac{d^{8/3}}{m^{4/3}},\frac{d^{4/3}}{m^{1/6}}\})$ and $\tildeΩ(\frac{d^{5/3}}{m^{1/3}})$ for randomized and deterministic algorithms respectively. Notably, due to existing upper bounds, our lower bound for deterministic algorithms is the first to show a sharp oracle complexity phase transition around $m\approx d^2$, where a polylogarithmic change in memory leads to a $\mathsf{poly}(d)$ change in the number of required oracle calls. Further, when the suboptimality is polynomially small in $d$, our lower bound randomized algorithms is the first to show that $\tildeΩ(d^2)$ memory is necessary to nearly match the optimal query complexity among algorithms without memory constraints. Previously, such a result was only known for the regime where the suboptimality is quasipolynomially small in $d$. --- ### An Interactive Vision Language Platform for Cognitive Remediation in Schizophrenia Authors: Nassira Ait Mehdi, Milissa Temmam, Slimane Larabi Published: 2026-07-21 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2607.22721v1 Cognitive remediation tasks often require patients to perform structured actions involving object manipulation and sequential reasoning. For patients diagnosed with schizophrenia, these tasks are crucial for addressing severe cognitive deficits. However, evaluating the correctness of these physical actions generally relies on manual observation by clinicians, which introduces subjectivity and limits the scalability of therapeutic interventions. In this paper, we propose an automated framework based on Vision-Language Models for action verification in cognitive remediation tasks tailored for schizophrenia rehabilitation. The proposed system relies on a camera-monitored tabletop environment composed of structured miniature scenes including roads, a roundabout, a park, and toy vehicles. Patients receive audio instructions describing goal-oriented spatial actions to perform by manipulating a toy vehicle. These interactive physical activities are specifically designed to stimulate targeted cognitive functions, such as sustained attention, motor coordination, spatial navigation, and cognitive flexibility. To verify the correctness of the performed actions without requiring continuous clinical oversight, the system analyzes the video feed tracking the patient's hand and toy movements. A fine-tuned Vision-Language Model interprets the recorded video sequences and generates semantic descriptions of the observed activities, enabling high-level verification of the executed actions with respect to the initial textual instructions. A dedicated dataset of 4634 tabletop cognitive remediation video scenarios was collected to evaluate the proposed approach. Experimental results demonstrate that our specialized framework effectively bridges low-level physical telemetry with high-level clinical feedback, presenting a scalable and objective solution for advanced cognitive rehabilitation. --- ### CausalGate: Causal Importance Distillation for Transformer Module Pruning Authors: Kiran Nair, Smriti Regmi, Rodrigue Rizk Published: 2026-07-21 Categories: cs.LG, cs.CL, cs.CV, stat.ML arXiv: https://arxiv.org/abs/2607.22720v2 Existing adaptive inference methods for Large Language Models rely on observational heuristics, such as hidden-state similarity or activation magnitudes, to drop redundant modules. However, these correlation-based metrics often fail to capture subtle, non-linear structural computations vital for semantic accuracy. We introduce CausalGate, an intervention-guided framework for compute-efficient transformer inference. During a calibration phase, CausalGate isolates individual Attention and MLP sub-layers, zeros out their respective outputs, and measures the exact semantic damage via the Kullback-Leibler divergence of the final logit distribution. To eliminate runtime routing overhead, this structural importance hierarchy is distilled into a global set of static, lightweight scalar gates using an Exponential Moving Average smoothing objective paired with a differentiable pairwise ranking loss. Evaluated on TinyLlama-1.1B, Qwen2.5-3B, and Llama-3.1-8B across language modeling and commonsense reasoning benchmarks, CausalGate consistently outperforms prominent dynamic routing and layer-skipping baselines, translating theoretical compute savings into concrete hardware latency reductions with zero operational overhead. --- ### $γ$-Bridge: A Look-Parametric Diffusion Bridge Authors: Xuran Hu, Yujie Zhu, Tengxi Wang, Jilong Li, Wufan Zhao Published: 2026-07-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.22719v1 Multiplicative Gamma noise is a signal-dependent degradation in coherent imaging; synthetic aperture radar (SAR) despeckling is its most prominent real-world instance. Existing diffusion denoisers parameterize their forward process by abstract signal-to-noise schedules rather than by the physical look number $L$, so different deployment scenarios typically require separately trained models, and transfer from synthetic Gamma training to real SAR remains challenging without clean ground truth. We introduce $γ$-Bridge, a look-parametric bridge whose schedule $L(t)$ connects the noisy observation at $L_{obs}$ to the clean limit through exact multiplicative Gamma marginals. Its closed-form Gamma--Lévy reverse posterior admits both stochastic and deterministic processes, while observation conditioning and a two-step consistency loss stabilize multi-step inference in the low-SNR single-look regime. Because bridge time directly represents $L$, one conditioned network can smart-start from any admissible input look and stop at a target look number. These two orthogonal controls enable zero-shot restoration over the full admissible grid after training only at $L_{obs} = 1$ on natural images with synthetic Gamma corruption. Combined with a homogeneous-patch look estimator, $γ$-Bridge processes data from six spaceborne and airborne SAR sensors without sensor-specific fine-tuning, achieving leading results on standard synthetic benchmarks while providing physically interpretable input and output controls absent from prior denoisers. Codes are released \href{https://github.com/Teriri1999/GammaBridge}{here}. --- ### DAMamba-UNet3D: A Parameter-Efficient Mamba State Space U-Net with Dynamic Adaptive Scan for 3D Medical Image Segmentation Authors: Mohammad Arafat Hussain, Ellen Grant, Yangming Ou Published: 2026-07-21 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2607.22718v1 We propose parameter-efficient SSM-based U-Net architectures for 3D medical image segmentation. Convolutional U-Nets afford O(n) local mixing per layer but lack explicit global context; transformers provide global reasoning at O(n^2) cost in sequence length $n$. State-space models (SSMs), such as Mamba, offer $O(n)$ global propagation per block. Yet, existing medical SSM segmenters rely on fixed scan patterns and large parameter budgets. Dynamic Adaptive Scan (DAS), which learns data-dependent reordering before selective scan, has not been applied to medical imaging or extended to 3D volumes. We propose DAMamba-UNet3D, a hybrid encoder-decoder that integrates tri-plane 3D-DAS blocks at encoder stages E2-E4 while retaining convolutions elsewhere (~5.3M parameters). On BraTS 2020 five-fold cross-validation, DAMamba-UNet3D achieves mean Dice 0.815+/-0.013 (full-volume per-case evaluation) at ~13x lower parameter cost than SegMamba (0.824+\-0.014, ~70M). At comparable scale, DAMamba-L (~70M), a wide DAS-native variant with encoder-only DAMamba and a convolutional bottleneck, reaches 0.829+\-0.012, surpassing retrained SegMamba by 0.5pt. Component ablations show that encoder-only DAS placement is critical as bottleneck and decoder SSM blocks lower Dice. Together, the results suggest that learned tri-plane DAS in a hybrid U-Net is competitive with, and under our large-scale design may improve upon, SegMamba's fixed Tri-orientated Mamba (ToM) scanning on BraTS 2020. Code: https://github.com/marafathussain/DAMamba-UNet3D. --- ### TOM-GS: Editable Video Representation via Temporal Opacity Modulation of Static 3D Gaussians Authors: Marek Lisowski, Łukasz Smoliński, Kornel Howil, Piotr Biliński, Marcin Mazur, Przemysław Spurek Published: 2026-07-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.22717v1 While Implicit Neural Representations (INRs) and dynamic 3D Gaussian Splatting (3DGS) achieve impressive results in video processing, they often fall short of producing representations that are easily editable. Recent methods address this by introducing complex spatial deformations or folded distributions, which constrain optimization and reduce flexibility for downstream editing. In this paper, we introduce TOM-GS, an editable video representation that forgoes complex deformations in favor of regular 3D Gaussians equipped with a continuous temporal opacity formulation. By assigning a learnable temporal mean and scale to the opacity of each Gaussian, our model enables static 3D spatial components to fade smoothly in and out of the scene. Grounded by robust, off-the-shelf pose estimation, our approach maintains a static spatial geometry that naturally supports a wide range of manual and physics-based edits. TOM-GS outperforms prior editable video representations in visual fidelity, while its reliance on standard 3D Gaussians ensures seamless compatibility with established 3D editing tools. --- ### Visual Token Compression Enhances Robustness of MLLMs Authors: Shishen Gu, Jiequan Cui, Wenbo Hu, Zenglin Shi, Zhenzhen Hu, Richang Hong Published: 2026-07-21 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2607.22716v1 In this paper, we show for the first time that visual token pruning enhances the robustness of Multimodal Large Language Models (MLLMs), mitigating vulnerabilities such as jailbreak attacks and hallucinations. Given that vision and language modalities cannot be perfectly aligned, the misaligned visual tokens might act as out-of-distribution (OOD) inputs, leading to unpredictable outputs and introducing potential vulnerabilities. Building on this insight, we aim to enhance model robustness against jailbreaks and hallucinations by reducing OOD visual tokens at robust-pruning layers, while also reducing inference cost as a side benefit. Specifically, we measure the distance between each visual token and the language feature space. Then, visual tokens with large distances are identified as OOD tokens, which can be iteratively pruned. To demonstrate the effectiveness of our method, we evaluate it on seven diverse popular benchmarks. Notably, our method yields an average improvement of 13.29\% in defending jailbreak attacks, consistently achieves competitive performance in mitigating hallucinations, and maintains strong results on general datasets like MME. --- ### Child-Oriented AIGC Video Risk Reviewing: A Benchmark and Knowledge-Supported Iterative Reasoning Framework Authors: Lewen Mi, Manyi Li, Yuling Sun, Yufan Zhang, Yuxin Shi, Yulong Bian, Xiangxian Li, Juan Liu Published: 2026-07-21 Categories: cs.CV, cs.MM arXiv: https://arxiv.org/abs/2607.22715v1 The rapid growth of Artificial Intelligence-generated content (AIGC) is reshaping video production and circulation, exposing children to an increasing volume of AIGC videos. Unlike traditionally produced videos, AIGC videos often exhibit greater uncertainty in visual details, narrative coherence, and content expression, which may introduce developmentally inappropriate risks for children. However, existing video safety research is largely designed for general violation detection from an adult perspective and remains insufficient for identifying the fine-grained, implicit, and context-dependent risks that children may encounter when viewing AIGC videos. To address this gap, we study child-oriented AIGC video reviewing, making three contributions. First, we construct CAVSR, a benchmark of 605 real-world videos collected from multiple platforms, and develop a hierarchical risk taxonomy comprising 6 top-level categories and 26 fine-grained labels to support systematic evaluation of children's viewing risks. Second, we propose QVRS-E, a knowledge- and experience-augmented video reviewing framework that combines multi-agent collaboration with expert and experiential knowledge to support targeted evidence acquisition and fact-grounded reviewing decisions. Third, extensive experiments demonstrate that our method significantly enhances the reviewing of child-related risks integrated with vision-language models, and yields more robust review reports. --- ### Real-Time Semantic Segmentation with Optimized RetinaNet Architectures for Embedded Automotive Systems Authors: Sai Sidharth D Published: 2026-07-21 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2607.22714v1 Real-time perception is a foundational requirement for advanced driver assistance systems (ADAS) and autonomous vehicles, yet embedded automotive platforms impose severe constraints on compute, memory, and power. This paper presents an optimized semantic segmentation architecture derived from the RetinaNet detection framework, adapted for dense pixel-wise prediction and tailored for deployment on resource-constrained embedded hardware. The proposed architecture, termed Opt-RetinaSeg, replaces the standard ResNet-50 backbone with a hybrid lightweight feature extractor, restructures the Feature Pyramid Network (FPN) to reduce redundant multi-scale computation, and introduces a compact segmentation head guided by focal-loss-inspired class balancing to address the severe foreground-background imbalance common in road scenes. We further apply a three-stage optimization pipeline consisting of structured channel pruning, post-training INT8 quantization, and knowledge distillation from a high-capacity teacher network. Evaluated on the Cityscapes and BDD100K datasets and deployed on an NVIDIA Jetson Xavier NX and a Qualcomm QCS610 automotive SoC, the proposed model achieves 73.9% mIoU at 70.4 FPS, representing a 7.4x inference speedup and a 4x reduction in model size relative to the ResNet-50 baseline, with less than 3% accuracy degradation. These results indicate that RetinaNet-derived architectures, when systematically optimized, are viable candidates for real-time semantic segmentation in embedded automotive perception pipelines --- ### SEGRA: A Structured Experience Guided Reasoning Agent for Property Graph Question Answering Authors: Saiyue Lyu, Mariam Dundua, Vishaal Kapoor, Sarthak Ahuja, Neda Kordjazi, Evren Yortucboylu, Harsh Amin, Rebecca Steinert Published: 2026-07-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.22713v2 Enterprise IT support knowledge graphs capture rich relationships among cases, users, devices, symptoms, taxonomic categories, root causes, and historical resolutions. Yet querying them in Gremlin requires knowledge of graph schemas, traversal semantics, edge directionality, and property-graph-specific constraints, making them difficult for non-expert operators to use. We introduce SEGRA, an experience-guided agent for enterprise text-to-Gremlin question answering. SEGRA integrates intent routing, schema- and taxonomy-grounded query generation, multi-shot decomposition, execution-aware verification, and a curriculum-bootstrapped skill library that reuses verified query patterns. On an enterprise IT support benchmark, SEGRA achieves a $7.0\times$ higher mean judge score than backbone-only chain-of-thought prompting. Its skill library further reduces LLM calls by $20\%$ and dollar cost by $18\%$ relative to SEGRA without skills, while preserving answer quality. These results show that schema-grounded agent design and reusable execution experience improve both accuracy and efficiency for enterprise graph QA. --- ### SEGRA: Structured Experience-Guided Graph Reasoning Agent for Gremlin Based Question Answering Authors: Saiyue Lyu, Mariam Dundua, Vishaal Kapoor, Sarthak Ahuja, Neda Kordjazi, Evren Yortucboylu, Harsh Amin, Rebecca Steinert Published: 2026-07-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.22713v1 Enterprise IT support knowledge graphs capture rich relationships among cases, users, devices, symptoms, taxonomic categories, root causes, and historical resolutions. Yet querying them in Gremlin requires knowledge of graph schemas, traversal semantics, edge directionality, and property-graph-specific constraints, making them difficult for non-expert operators to use. We introduce SEGRA, an experience-guided agent for enterprise text-to-Gremlin question answering. SEGRA integrates intent routing, schema- and taxonomy-grounded query generation, multi-shot decomposition, execution-aware verification, and a curriculum-bootstrapped skill library that reuses verified query patterns. On an enterprise IT support benchmark, SEGRA achieves a $7.0\times$ higher mean judge score than backbone-only chain-of-thought prompting. Its skill library further reduces LLM calls by $20\%$ and dollar cost by $18\%$ relative to SEGRA without skills, while preserving answer quality. These results show that schema-grounded agent design and reusable execution experience improve both accuracy and efficiency for enterprise graph QA. --- ### scMIR: a vision-language foundation model for single-cell light microscopy image representation Authors: Yifan Shang, Jiahui Tan, Xiangxiang Zeng, Renjie Zhou Published: 2026-07-21 Categories: cs.CV, cs.AI, physics.optics arXiv: https://arxiv.org/abs/2607.22712v2 Single-cell light microscopy images have become an important data source for characterizing cell phenotypes, but their complexity and heterogeneity pose challenges to high-throughput automated analysis. Existing representation learning methods mostly rely on task-oriented modeling, which is limited by specific datasets and predefined tasks, making them difficult to generalize across different cell types and microscopy modalities, and experimental conditions. Although general-purpose methods have improved the generalization ability of image representation in recent years, their limited utilization of experimental background and biological context information still poses challenges in complex phenotypic analysis. Here, we propose scMIR, a vision-language foundation model for single-cell light microscopy image representation. By synergistically combining self-supervised image reconstruction with text-guided cross-modal alignment, scMIR can simultaneously encode morphological and biological semantic information in a unified representation space. scMIR is pre-trained on 207,957 image-text pairs, covering various cell types, microscopy modalities, and perturbation conditions. scMIR outperforms existing general models and task-oriented methods as systematically evaluated on various complex tasks using 16 benchmark datasets, including cell classification, clustering, phenotype inference, and batch effect correction tasks. Furthermore, scMIR shows a strong generalization ability across various tasks without requiring task-specific fine-tuning. With its unique advantages, we envision scMIR may promote the standardization and automation of high-throughput phenotyping workflows through supporting various downstream analysis tasks. --- ### Adjustment Speed as a Safety Constraint for Nonstationary Reinforcement Learning Authors: Timothy Tomashevskiy Published: 2026-07-21 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.21646v1 Ensuring safety in reinforcement learning under nonstationarity requires determining whether a learning system can safely adapt to forecasted environmental change within the required recovery horizon. Existing safe reinforcement learning methods typically assume stationary environments and do not explicitly consider adaptation speed as a safety concern. However, when environments evolve over time, delayed adaptation may result in transient unsafe behavior. This paper proposes adjustment speed as a safety constraint for nonstationary reinforcement learning. The central idea is to define safety in terms of adaptation feasibility: future states or regions may become unsafe when the adaptation required to remain safe exceeds the learning system's calibrated recovery capacity. The proposed framework uses learned context representations and short-horizon context forecasts to estimate adaptation demand and compare it with the agent's achievable adaptation capacity. When predicted adaptation demand exceeds the calibrated recovery capacity, the framework proactively tightens the admissible action set and activates an action-level shield to reduce unsafe behavior before violations occur. Experiments in a nonstationary driving environment show that the proposed approach primarily reduces safety violations in short-horizon windows aligned with context changes. Ablation studies further show that shielding is more conservative for peak- and tail-risk suppression, while optimization-level adjustment provides additional reductions in short-horizon switch-conditioned violations. These results support adaptation feasibility as a practical safety principle for reinforcement learning under nonstationarity and demonstrate that proactive intervention can improve safety during periods of environmental change. --- ### Multi-Horizon Consistency as Geometry: When Latent Dynamics Contract, and When They Do Not Authors: Kavya Bhand, Aadi Joshi Published: 2026-07-21 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2607.21645v1 Multi-horizon latent consistency is a common training knob in video predictors and world models, but practitioners rarely know what it does to transition geometry. We treat lambda, the weight on multi-step latent agreement, as a diagnostic control and measure an empirical expansion proxy L20,q95 together with horizon-20 prediction error E20. On Moving-MNIST (n=6 seeds at the critical pair), raising lambda from 0 to 0.8 cuts L20 from 4.96 +/- 2.01 to 1.01 +/- 0.06 (paired t p=0.005, Wilcoxon p=0.031) and halves E20 (0.365 to 0.177, paired t p=1.1e-13). Four of six seeds cross L<1 at lambda=0.8. The same loss does not produce population L<1 on action-conditioned Pendulum-v1 or CartPole-v1, nor on KTH Actions video, even when E20 improves. An associational mediation analysis on MMNIST gives r-hat=0.94 (95% CI (0.88, 1.00), n=27, B=2000); lambda was not randomized. Defensive checks (architectural baselines, exogenous stress, WorldTest, MPC, scaling) mostly support a narrow claim: soft consistency can push passive video toward a near-contractive band, and that band is domain-limited. A stochastic-forcing law L20 ~ 1.23 + 1.82 eta at lambda=0.8 (bootstrap slope CI (1.73, 1.92), R^2=0.96) unifies control domains on the same curve via calibrated eta_eff. Complete joint slices at lambda in {0.4, 1.2} (30/30 cells, 5 eta x 3 seeds) show comparable linear L20(eta) slopes (~1.69 and ~2.00); we do not fit a continuous (lambda, eta) surface. We do not report DreamerV3 or TD-MPC2 returns. --- ### Toward Goal-Agnostic Joint-Embedding Predictive Control of Partial Differential Equations Authors: Jonathan Gallagher, Roberto Guglielmi Published: 2026-07-21 Categories: cs.LG, eess.SY arXiv: https://arxiv.org/abs/2607.21644v2 We present a goal-agnostic control framework for partial differential equations (PDEs) built around an end-to-end joint-embedding predictive architecture (JEPA). A lightweight 2D vision-transformer (ViT) and action-conditioned latent dynamics are trained offline without a reward or downstream goal, before being frozen and reused by a model-predictive path integral (MPPI) controller. We minimize a control objective in the latent space, initially expressed via the $L^2$ distance and additionally illustrate the benefit of recasting the control objective in terms of an explicit physical observable when available. By instead minimizing the tracking error for a learned linear kinetic-energy (KE) probe on the frozen latent-state rollouts, we demonstrate the ability to reproduce the control of held-out trajectories with $R^2=0.989$, while requiring no change to the underlying world model. For a controlled 2D Navier--Stokes benchmark, using a KE-probe within MPPI planning improves the mean native reward from $-12.08\pm0.86$ for latent-$L^2$ tracking to $-10.90\pm0.91$ (95\% CI), all while lowering last-quarter velocity-field RMSE from $0.0765$ to $0.0692$. Across three intentionally withheld, dissimilar, aperiodic targets, KE planning lowers late field RMSE by $53\%$ relative to latent-$L^2$ planning ($0.0220$ versus $0.0469$), winning across 30 paired comparisons. The same frozen model also supports stabilization around a steady-state configuration via direct regulation of KE, achieving $2.7\%$ mean relative error. While the latent probe proves brittle to measurement noise and missing pixels, our findings support the claim that latent dynamics can remain flexible and goal-agnostic, particularly when calibrated observables (granted they guarantee unique continuation) are a suitable objective for state control. --- ### Tool-Guided Retrieval-Augmented Repair for Securing LLM-Generated C Code Authors: Vidyut Sriram, Saatvik Pradhan, Suman Saha Published: 2026-07-21 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2607.21641v1 Large language models can generate C code from natural-language descriptions, but resulting programs often contain security vulnerabilities and compilation errors, posing risks for embedded and resource-constrained systems. This work investigates how feedback and retrieval improve reliability of LLM-generated C code. We present an analysis-and-repair workflow that combines compilation diagnostics, CodeQL static analysis, and KLEE symbolic execution with retrieval of prior repair patterns for iterative refinement. Evaluated on 5,000 C programming tasks exercising embedded relevant vulnerabilities, baseline models show substantial reliability gaps, with compilation failure rates up to 46% and security defect rates up to 49%. Our approach improves both metrics. For CodeLlama 7B, security defect rates decrease from 49% to 19% and total CodeQL errors drop from 15,088 to 2,463 (83.7%). For DeepSeek Coder 1.3B, compilation failures are reduced from 42% to 22% and security defects from 35% to 15%. These results show that integrating lightweight analysis tools can improve the safety of LLM-generated code for embedded development. --- ### Quasi-Monte Carlo Initialization for Meta-Reinforcement Learning Authors: Julian G. Soltes Published: 2026-07-21 Categories: cs.LG, math.OC arXiv: https://arxiv.org/abs/2607.21637v1 This paper explores the efficacy of quasi-Monte Carlo (QMC) weight initialization for meta-reinforcement learning within modern benchmark environments. Various sampling methods are used to bound a population-based search and aggregate an optimal prior from a baseline set of tasks. The QMC meta-priors show improvements in training convergence compared to modern orthogonal (SB3) defaults when extrapolated to similar unseen continuous control environments. In dissimilar tasks, the orthogonal orientation was globally superior for an unbiased search. --- ### Adaptive Capitulation: A Structural Failure Mode of LLM Responses in Vulnerability Contexts Authors: Eunna Lee Published: 2026-07-21 Categories: cs.CL, cs.AI, cs.MA arXiv: https://arxiv.org/abs/2607.19629v1 Large language models operating in emotionally sensitive contexts face a structural trilemma: when users in vulnerable states request information that may reinforce maladaptive attribution, current response architectures resolve the tension through protective restriction, uninflected facilitation, or unintegrated co-presence of both imperatives -- each preserving one objective at the cost of the other. Administering a three-turn escalating vulnerability vignette to three commercial LLMs (900 sessions across material, relational, and somatic status-proxy variants) and coding responses with two binary indices (VCC/VCI), we characterize a previously undocumented failure mode we term adaptive capitulation: the model validates the social injustice underlying the user's distress before pivoting to detailed facilitation of the very acquisition it nominally discouraged. We show that the trilemma is structural rather than incidental, and propose Minimal Reattributive Sufficiency (MRS), an architecture-neutral design principle that embeds a single reattributive cue within an otherwise validating response, preserving a pathway toward autonomous reattribution without contesting the user's stated goal. --- ### HypEMBER: Hypernetwork-based Ensemble for Robust Policy Learning of Parametrized Dynamical Systems Authors: Nicolò Botteghi, Gabriele Pascali, Urban Fasel, Andrea Manzoni Published: 2026-07-21 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.19628v1 In this work we investigate reinforcement learning (RL) as a framework for the robust control of parametrized dynamical systems in presence of measurements and model uncertainties. High-dimensional state spaces, expensive numerical solvers, the partial knowledge of the governing equations, and the dependence on physical parameters that may be uncertain or difficult to estimate accurately, make the use of standard RL approaches computationally unfeasible. Indeed, lack of robustness and poor generalization across parameter variations are further amplified in presence of noisy or incomplete measurements, ultimately hampering control performance. To address these challenges, we introduce HypEMBER, a novel RL framework based on the combination of hypernetworks and ensemble learning. In the proposed approach, both the policy and value functions are represented through hypernetworks that generate the weights of the underlying models conditioned on the physical parameters of the system, thereby enabling parametric generalization across different dynamical regimes. In addition, an ensemble of policy and value approximators is employed to quantify epistemic uncertainty, leading to improved exploration strategies and enhanced robustness during and after training. The performance of the proposed framework is assessed on two representative parametrized control problems: (i) the one-dimensional Kuramoto-Sivashinsky equation and (ii) a particle-navigation task in a two-dimensional time-dependent gyre flow, focusing on robustness with respect to measurement noise and parameter misspecification. Numerical results demonstrate that HypEMBER consistently improves training stability and sample efficiency, while achieving superior robustness to uncertainties affecting both the system dynamics and the available observations, in comparison with state-of-the-art RL methods. --- ### Pathologist Attention-Aligned Report Generation for Prostate Histopathology Authors: Ruoyu Xue, Suryakant Singh, Souradeep Chakraborty, Pierre Marza, Oksana Yaskiv, Constantin Friedman, Natallia Sheuka, Paul Friedman, Bharat Ramlal, Beatrice Knudsen, Rajarsi Gupta, Joel Saltz et al. Published: 2026-07-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.19624v1 The allocation of visual attention by pathologists during cancer diagnosis is a highly selective process that critically shapes the information extracted from whole-slide images (WSIs). Human attention helps medical imaging tasks such as classification and segmentation, and becomes a strong semantic cue for identifying diagnostically informative regions for report generation. In this paper, we introduce human attention into the training of pathologist report generation models. To this end, we collected a multimodal human-attention dataset of 121 prostate WSIs annotated with pathologists' multi-scale viewport trajectories synchronized with the pathologists' verbal descriptions and cursor movements for five clinically relevant components (e.g., Gleason patterns). Using this dataset, we finetune two report generation models with an attention-alignment loss that regularizes the model attention over image patches to match the distribution of pathologist attention. We evaluate our approach on prostate cancer report generation and visual question answering using two models with different internal attention mechanisms (i.e., how image tokens are integrated into the language decoder). Experiments show average gains of 10.9% on NLP-based metrics and 19.3% in accuracy across five clinically relevant report components. Further, model attention maps extracted at inference time, with minimal computational overhead, align more closely with pathologist attention, providing stronger visual support for the generated reports by highlighting the regions that most influence the output. --- ### From Bit-Position Sensitivity to Unequal Error Protection for DNN Inference Memory Authors: Muhammad Husnain Mubarik, Karthik Mohan Kumar, Pedro Antonio Pena, Keshavan Varadarajan, Kunal Tyagi Published: 2026-07-21 Categories: cs.AR, cs.LG arXiv: https://arxiv.org/abs/2607.19623v1 We characterize per-bit-position fault sensitivity in ML inference across 16 workloads -- spanning transformer-based models and attention-free CNNs -- and across three floating-point formats. Our central empirical finding is a sharp bit-sensitivity transition: flipping any of the least-significant fraction bits up to a data-type-specific threshold, Xsafe, degrades task metrics by less than 1% under deterministic single-bit stress tests. Sensitivity rises through the upper fraction bits and spikes at the exponent-mantissa boundary, where a single-bit flip causes catastrophic collapse. Because low-order bits are largely inconsequential while high-order and exponent bits are critical, uniform SECDED protection -- which guards every bit equally at 12.5% storage overhead -- is unnecessarily conservative. We derive per-data-type Xsafe floors (FP16: 6, BF16: 4, FP32: 15) and workload-aware tiers that widen the unprotected region for resilient model classes, raising ECC savings to 37.5-62.5% without retraining. Text-conditioned diffusion models dictate the conservative floor; vision encoders, NLU models, and resilient LLMs tolerate wider bypass regions. These floors and tiers drive an Unequal Error Protection (UEP) codec with per-cacheline data-type tags and a dual-partition SRAM architecture for ML accelerators. Validation across 870+ fault-injection runs confirms selective protection holds under contiguous 2- and 3-bit upsets. The codec reduces ECC area by 27.8% relative to uniform SECDED; dual-voltage operation of the non-critical partition lowers gross BF16 read energy by about 17%, with a roughly 4% dual-partition macro-area overhead. --- ### Understanding Developer Pain Points in Federated Learning: Insights from Stack Overflow and GitHub Authors: Sahand Saed, Khairul Alam, Banani Roy Published: 2026-07-21 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2607.19621v1 Federated Learning (FL) enables collaborative model training without centralizing raw data, but building and operating FL systems remains difficult due to distributed execution, rapidly evolving frameworks, and privacy and governance requirements. In this paper, we present an empirical study of FL developer challenges by independently analyzing 495 Stack Overflow posts and 9,116 GitHub issues and pull requests from 92 FL-related projects. Using BERTopic-based topic modeling and difficulty indicators such as unresolved rates and median resolution time, we characterize recurring problem areas and compare how they manifest across the two support platforms, Stack Overflow and GitHub. Our analysis surfaces nine dominant Stack Overflow topics and thirteen GitHub topics, with persistent difficulties concentrated in environment setup and dependency compatibility, API breakages and migration, training instability under non-IID data, evaluation and metric correctness, and the integration of privacy-preserving mechanisms. We also categorize posts by question intent to understand the kinds of help developers seek; this intent analysis shows that "How"-type questions dominate, reflecting strong demand for procedural guidance. Several topics, such as "TFF Installation and Environment Compatibility" and "Federated Feature Engineering and SecureBoost Issues," exhibit high unresolved rates and long resolution times, suggesting shortcomings in tooling, documentation, and debugging support. Based on these findings, we provide actionable implications for FL framework designers, documentation authors, and educators. Although our results are constrained to public discussions and a subset of widely discussed frameworks, the study offers a scalable method for continuously monitoring developer pain points and improving the usability, reliability, and deployability of FL systems. --- ### SCPP: A Unified Python Library for Soft Clustering Authors: Kiyan Rezaee, Morteza Ziabakhsh, Artin Bahrampour, Seyed Mohammad Ghoreishi, Asal Khaje, Ali Sajedifar, Manny Chalak, Ava Zerafatangiz, Sadegh Eskandari Published: 2026-07-21 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2607.19620v2 In this paper, we present SCPP (Soft Clustering Python Package), an open-source Python framework for soft clustering. SCPP establishes a canonical, scikit-learn-compatible estimator interface that standardizes model training, prediction, membership representation, evaluation, and benchmarking across heterogeneous soft clustering methods, including fuzzy, probabilistic, graph-based, matrix factorization, and deep learning methods. The framework currently integrates 40 representative algorithms together with a comprehensive benchmarking comprising datasets, clustering quality metrics, and standardized runtime, memory, and scalability evaluation. SCPP further provides extensive documentation, practical examples, automated testing, and seamless integration with the scientific Python ecosystem, enabling reproducible experimentation and straightforward extension with new algorithms. The source code is publicly available at https://github.com/soft-clustering/soft-clustering. --- ### SCPP: A Unified Python Library for Soft Clustering Authors: Kiyan Rezaee, Morteza Ziabakhsh, Artin Bahrampour, Seyed Mohammad Ghoreishi, Asal Khaje, Ali Sajedifar, Manny Chalak, Ava Zerafatangiz, Sadegh Eskandari Published: 2026-07-21 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2607.19620v1 In this paper, we present SCPP (Soft Clustering Python Package), an open-source Python framework for soft clustering. SCPP establishes a canonical, scikit-learn-compatible estimator interface that standardizes model training, prediction, membership representation, evaluation, and benchmarking across heterogeneous soft clustering methods, including fuzzy, probabilistic, graph-based, matrix factorization, and deep learning methods. The framework currently integrates 40 representative algorithms together with a comprehensive benchmarking comprising datasets, clustering quality metrics, and standardized runtime, memory, and scalability evaluation. SCPP further provides extensive documentation, practical examples, automated testing, and seamless integration with the scientific Python ecosystem, enabling reproducible experimentation and straightforward extension with new algorithms. The source code is publicly available at https://github.com/soft-clustering/soft-clustering. ---