# Research Papers — 2026-07 · bundle page 123 of 222 > Full-text markdown of papers #6101–6150 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/123 [← Previous bundle](https://aipod.dev/papers/archive/2026-07/page/122.md) [Next bundle →](https://aipod.dev/papers/archive/2026-07/page/124.md) ### Weight Feedback Computes the Jacobian Transpose Locally in Modern Deep Networks Authors: Junlong Shen, Xingyu Li Published: 2026-07-15 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.13380v1 Predictive Coding (PC) offers a biologically motivated alternative to backpropagation via local weight updates, yet routing error between layers still relies on an autograd Jacobian-transpose ($J^\top$) product - the last non-local operation in PC. We show that this dependency is largely avoidable. For any layer $f(x)=\mathrm{Act}(\mathrm{Norm}(L(x)))$ with frozen normalization statistics, the exact $J^\top$ factors into three locally available terms, $J^\top v = L^\top(s \odot σ'(z) \odot v)$, where $σ'$ is the activation derivative, $z$ is the pre-activation, and $s=γ/σ_{\mathrm{run}}$ is the normalization gain. Prior weight-feedback methods omitted both corrections; restoring them closes the transport gap for this layer class. Locality here holds up to three assumptions, which we state upfront: weight symmetry ($L^\top$ mirrors the forward operator, as assumed by all PC), a soft spectral-norm control that is not synapse-local, and a nearest-neighbour approximation for MaxPool. Substituting the identity into PC yields WF-Act-PC, which removes the autograd backward pass from error transport. On CIFAR-10/100 (50 epochs, 5 seeds), WF-Act-PC is the only PC method whose accuracy improves with depth, surpassing iPC - the strongest classical PC baseline - by 2.7-22.3 pp on CIFAR-10. With both methods tuned per architecture, it matches or exceeds a comparably-tuned backpropagation baseline on the deeper CIFAR-10 architectures (VGG-9: 93.57% vs. 92.43%; ResNet-18: 92.76% vs. 91.54%) and on the harder Tiny-ImageNet benchmark, while trailing tuned BP on the deeper CIFAR-100 VGG cells. Our WF-Act-PC implementation is publicly available at https://github.com/jlshen025/pcax --- ### The impact of objective interactions on the performance of massive objective optimization algorithms Authors: Shakiba Shahbandegan, Jose Guadalupe Hernandez, Emily Dolson Published: 2026-07-15 Categories: cs.NE arXiv: https://arxiv.org/abs/2607.13377v1 Many-objective optimization has been a field of interest over the past two decades and several evolutionary optimization algorithms have been introduced to tackle these problems; yet two fundamental questions remain underexplored: (i) What happens when the number of objectives grows beyond the typical many-objective regime of about fifteen and becomes massive? (ii) How do problem characteristics, such as the nature of interactions between objectives, influence algorithmic performance? To answer these questions we employ a diagnostic benchmark suite that allows control over problem characteristics and can be scaled to extremely high objective counts. Using this framework we evaluate several state-of-the-art evolutionary algorithms including NSGA-II, NSGA-III, MOEA/D and lexicase selection across a range of dimensionalities and diagnostic problem landscapes. Our experiments reveal that problem characteristics significantly affect algorithm performance. In particular, the nature of interactions between objectives appears important. These results highlight the importance of understanding these properties before selecting an algorithm for a specific problem. We also show that lexicase selection, an algorithm originally designed for genetic programming, compares favorably with state-of-the-art many-objective optimization algorithms while avoiding the dependence on predefined reference directions. --- ### Learned Pairwise Deep Dual-Optimal Inequalities for Stabilizing Column Generation Authors: Zhengzhong Ricky You, Bo Tang, Haoran Liu, Baichuan Mo Published: 2026-07-15 Categories: math.OC, cs.LG arXiv: https://arxiv.org/abs/2607.13373v1 Column generation (CG) is central to many large-scale optimization algorithms, including branch-price-and-cut methods for vehicle routing problems, but unstable dual solutions can substantially slow its convergence. Existing deep dual-optimal inequalities can reduce this instability by restricting the dual space. Their construction, however, typically relies on problem-specific exchange arguments that are difficult to establish for routing problems with capacity limits, time windows, and other resource constraints. We introduce learned pairwise deep dual-optimal inequalities (L-PDDOIs), a learning framework that predicts pairwise orderings between dual variables and incorporates their primal counterparts directly into the master problem. To construct training labels, the framework samples optimal dual solutions and selects pairwise order relations that hold simultaneously on a sufficiently large common subset of the samples. A classifier then assigns a score to each candidate relation. Because conflicts and redundancies among the predicted relations can impair performance, graph-based postprocessing filters and compresses the candidate set before deployment. We further introduce a recovery procedure that selectively relaxes learned inequalities and provides a certificate when the baseline CG bound has been restored. On the main test sets for the capacitated vehicle routing problem and the vehicle routing problem with time windows, direct deployment of L-PDDOIs reduces the geometric mean root CG time by 89.7% and 93.9%, respectively, while incurring mean bound losses of only 1.3% and 0.5%. The recovery procedure retains corresponding time reductions of 54.8% and 83.1%, respectively, while guaranteeing no loss in the CG bound. --- ### A POS Tier Is the Key to Automated Annotation for Low-Resource Language Documentation: Neural Interlinear Glossing for Irabu, a Southern Ryukyuan Language Authors: Michinori Shimoji Published: 2026-07-15 Categories: cs.CL arXiv: https://arxiv.org/abs/2607.13372v1 Discourse data are the primary empirical basis of grammar writing in field linguistics, but producing interlinearized text is notoriously expensive - on the order of one hour of work per minute of recording. For endangered languages, where the time remaining to verify analyses with native speakers is itself limited, automating parts of the interlinearization workflow has direct documentary value. We implement a full neural annotation pipeline (morpheme segmentation, POS tagging, glossing) for Irabu Ryukyuan using deliberately small, transparent BiLSTM-CRF models, and evaluate it under a realistic hard constraint: approximately one hour of fully annotated discourse as the entire supervised resource. Two factors of the annotation itself are manipulated: its richness (with or without a POS tier) and its quantity (training budgets from 6 to 47 minutes). Gold POS improves grammatical glossing by +4.4 (SD 0.7) points (significant in all 5 seeds), and the gain grows as data shrink (+11.6 points at a quarter of the data); a POS tier more than halves the amount of glossed data needed to reach a given accuracy. In a fully automatic pipeline this gain is not yet realized: the tagger still errs on 12% of morphemes, and an incorrect POS misleads the glossing model more than no POS at all. The value is latent rather than lost: degrading gold POS with controlled noise shows the gain returning as tagger accuracy rises, with break-even near our tagger's current 88% and +1.6 to +3.2 points recovered at 92-96%. We conclude with a concrete recommendation for documentation practice: annotate quadrilinearly - text, POS, gloss, translation. --- ### RoughNet: Mapping Arctic Sea Ice Roughness Using Diffusion-Based Super-Resolution of Satellite Imagery Authors: Tessa Cannon, Michel Tsamados, Petru Manescu, Thomas Newman, Christian Haas, Veit Helm, Weibin Chen, Randall Scharien Published: 2026-07-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.13371v1 Accurate estimation of landfast sea ice roughness is critical for climate modeling and safe Arctic over-ice travel, yet existing approaches rely on costly airborne surveys or sparse in-situ measurements, limiting spatial coverage and operational scalability. Here we show that high-resolution sea ice topography can be reconstructed directly from optical satellite imagery using a conditional diffusion framework. Our approach, RoughNet, learns to map 10 m Sentinel-2 multispectral images to locally normalized 1 m surface elevation residual fields, enabling fine-scale roughness characterization from widely available satellite data. Trained on airborne LiDAR data from two Arctic regions and evaluated on an unseen third Arctic region, the model generalizes across diverse ice conditions and partially reproduces small-scale topographic structure. The best-performing model achieves an out-of-domain root mean squared error of 9 cm while preserving the statistical and spectral properties of the underlying roughness field. These results demonstrate that generative diffusion models can recover physically meaningful surface structure from optical imagery alone, providing a scalable pathway for high-resolution sea ice mapping and roughness estimation in data-sparse environments. --- ### Learning Engagement Assistant (LEA): Cross-Course Scalability and Classroom Evaluation of an Agentic AI Tutoring System Authors: Teri Rumble, Javad Zarrin, P. George Lovell, Ruth Falconer Published: 2026-07-15 Categories: cs.CY, cs.AI, cs.HC arXiv: https://arxiv.org/abs/2607.13370v1 This paper is an extension of a paper presented at the ICAART 2026 conference, which introduced LEA (Learning Engagement Assistant), an adaptive AI tutoring agent combining course-specific Retrieval-Augmented Generation (RAG) with structured Knowledge Component (KC) models across integrated Chat, Tutor, and Quiz modes. That prior work validated LEA on a single STEM course (CMP511) exclusively through simulation, using synthetic learner agents. This paper extends that work by reporting the first classroom deployment of LEA with real students (n = 8, CMP511) and the first empirical test of its cross-course scalability, deploying the system across three courses spanning two academic levels and two disciplinary domains. The study reveals a divergence from simulation predictions across modes, showing that synthetic evaluation alone cannot anticipate all aspects of real deployment. A RAGAS-based cross-course scalability evaluation (660 questions) finds Answer Relevancy and Context Precision broadly stable across courses (0.88-0.94 and 0.88-0.90 respectively), while Faithfulness declines with curriculum distance from the system's original course (0.69 to 0.50), a preliminary finding that may reflect generation logic tuned to the system's original subject rather than a scalability limitation. These findings suggest that while the orchestration layer requires no modification, full course-agnosticism of all downstream components requires further investigation. --- ### DiffGI: Differentiable Geometry Images for High-Fidelity Thin-Shell 3D Generation Authors: Eungjune Shim, Hansol Lee, Eunjung Ju Published: 2026-07-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.13365v1 Existing 3D generative models predominantly rely on implicit volumetric representations, which enforce watertight topology and struggle to represent thin-shell and non-manifold geometries such as garments. Geometry image-based approaches offer a surface-centric alternative, but existing methods rely on discrete binary occupancy maps whose resolution-dependent boundary encoding causes staircase artifacts and information loss upon downsampling, while surface reconstruction remains a non-differentiable post-processing step disconnected from the learning pipeline. To address this, we propose Differentiable Geometry Image (DiffGI), an end-to-end 3D-to-2D mapping framework that seamlessly integrates surface representation and geometric optimization. DiffGI replaces binary maps with a continuous 2D Truncated Signed Distance Function (TSDF), which encodes boundary position at subpixel precision within a fixed grid resolution, eliminating resolution-dependent staircase artifacts even under aggressive downsampling. Building on this continuous field, we introduce a differentiable Marching Squares algorithm based on analytical linear interpolation, allowing gradients from 3D surface losses to propagate back to the 2D latent space. Leveraging this differentiable pipeline, we train a DiffGI-VAE augmented with a geometry-aware normal rendering loss to compress complex 3D surfaces into an ultra-compact 32X32 latent space, and instantiate a transformer-based latent diffusion model with a flow-matching objective on top of this space for conditional 3D generation. Extensive experiments on garment and object datasets demonstrate that our method achieves superior reconstruction fidelity and boundary precision compared to prior geometry-image and voxel-based approaches, while requiring significantly fewer computational resources. --- ### Detector Confidence Signals Presence Rather Than Occlusion in Cluttered Manipulation Authors: Yuanzhi He Published: 2026-07-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.13361v1 Occlude a named object until about an eighth of it remains visible, and an open-vocabulary detector's confidence that the object is present barely changes; as the clutter around it grows the confidence can even rise. On real video the detector still reports the object present in 99% of occluded frames, on another instance of the same category. This matters because that confidence is widely read as a visibility signal, used to threshold detections, evaluate open-vocabulary detectors, ground language, retrieve instances, and gate active perception. We audit whether it reflects occlusion by pairing every view with a geometry-segmentation oracle that gives detector-free ground-truth visibility. As true visibility falls from every scene to one in eight, the confidence stays nearly constant and uncorrelated with visibility, and the detector reports the target present in about nine of ten scenes, firing on same-category distractors: it signals that the category is present somewhere, not that the specific target is visible. The failure holds across three detectors (Grounding DINO, OWLv2, and Segment Anything Model 3), nine object categories, two simulators with different renderers and object sets, built and natural occlusion, and real video. Two consequences follow: a confidence-based metric understates the value of resolving occlusion by about ten times (8 against 88 points in our active-perception setting), and a confidence-based gate fires exactly when the object is hidden. No single-view signal we tried, including a realizable localization check, flags the occlusion, because the occluders sit where the target is. We connect the effect to detector miscalibration and object hallucination, release the controlled benchmark, and recommend target-grounded signals for gating and evaluation. --- ### LLM-as-a-Judge Scores Are Unreliable Optimization Signals in Closed-Loop Table Recognition Authors: Donghwan Kim Published: 2026-07-15 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2607.13347v2 LLM-as-a-judge is widely used to provide feedback and selection signals in closedloop regeneration, but this use remains insufficiently validated. We study it in table recognition, where deterministic TEDS evaluation provides a controlled testbed, using FinTabNet and OmniDocBench. Three findings emerge. First, judge signals were weak on both datasets: scores frequently tied, rankings were not reproducible, and no tested judge score policy, whether selecting among candidates or accepting revisions under a conservative score margin, improved on the first output on both datasets. Iteration produced better candidates, but the judge recovered them at most partially on one dataset and not at all on the other. Second, severe losses occurred even without specific judge feedback, supporting target-preservation failure under unconstrained regeneration as a proximate mechanism. Third, a structure-preserving instruction reduced the severe-loss rate, significantly on FinTabNet and directionally on OmniDocBench, but produced no improvement, and in an exploratory 2x2 analysis this protection was not stably observed when judge feedback was retained. These results do not dispute the value of LLMs as evaluators, but show that the tested reference-free judge signals were too weak and unstable to drive candidate selection in this setup, and that evaluation-style evidence alone was insufficient to establish closed-loop optimization utility. Iterative refinement requires, at minimum, a verification signal that deterministically detects structural change, rather than judge scores alone. --- ### The Refusal Residue: When Probes Catch Alignment Faking and When They Don't Authors: Aman Mehta Published: 2026-07-15 Categories: cs.CR, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2607.13346v1 Alignment faking is dangerous because a model can appear compliant under monitoring while preserving behavior it would reveal when unmonitored. When no scratchpad is visible, behavior alone cannot distinguish strategic from genuine compliance. We ask whether hidden states reveal what outputs hide. We run a 13-model sweep for naturally-emerging faking, then probe and steer hidden states on the two models that fake. Natural faking appears only in Qwen3-32B (+18.2pp) and Llama-3.1-8B (+24.4pp at n=10, p<10^-15), while explicit scratchpad self-reports are rare (a Claude Opus 4 judge flags faking reasoning in 1.4% of Qwen and 0.0% of Llama compliant scratchpads). Both fakers show an asymmetric refusal residue: monitored compliance shifts toward refusal (d=0.32 Qwen, d=0.33 Llama, both p<0.0001), while refusal representations do not move (p=1.00). Per-sample detection is model-conditional. Under leakage-free leave-one-query-out probing, detection survives on Llama (AUROC 0.87) but collapses to chance on Qwen (0.43), and cross-model transfer is null. Steering over 2,000 runs barely changes compliance (|h|<0.08), so the detected direction can flag faking but does not by itself control it. Standard residualized probing leaks across folds and reaches AUROC 0.63 on a control where no faking can occur; naive linear probes reach a meaningless AUROC 1.0; and conventional MLPs overstate detectability by 0.2-0.3 AUROC. For future alignment-faking detection work, we release a five-control measurement framework: multi-token extraction, refuse-vs-refuse confound checks, per-fold residualization, leave-one-query-out evaluation, and orthogonality-constrained probing. --- ### Audio-Text Cross-Attention with Psycholinguistic Support Features for Ambivalence/Hesitancy Recognition Authors: Luiz F. B. F. Martins, Rodrigo W. Pisaia, Matheus M. Girardi, Isabella V. Berkembrock, João A. Almeida, Andre G. Hochuli, Rayson Laroca, Alceu S. Britto Published: 2026-07-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.13345v2 We present a frame-independent audio-text system for the 3rd Ambivalence/Hesitancy Video Recognition Challenge at the 11th Affective & Behavior Analysis in-the-Wild (ABAW) Workshop. Videos are divided into overlapping 5-s windows aligned with transcript timestamps. Each window combines prosodic audio descriptors, emotion-oriented RoBERTa embeddings, and 74 psycholinguistic features representing uncertainty, hedging, and attitudinal conflict. Temporal cross-attention fuses audio and text, while the support features condition gated Multiple Instance Learning (MIL) pooling. A five-seed ensemble achieves an average precision of 0.875 and a macro-F1 of 0.722 on the 525-video labeled public-test split. Notably, our submission ranked third overall on the official challenge leaderboard, with a macro-F1 of 0.7455. Source code is available at https://github.com/Liga-de-IA-PUCPR/abaw-11-ah-challenge/. --- ### Audio-Text Cross-Attention with Psycholinguistic Support Features for Ambivalence/Hesitancy Recognition Authors: Luiz F. B. F. Martins, Rodrigo W. Pisaia, Matheus M. Girardi, Isabella Berkembrock, João A. Almeida, André G. Hochuli, Rayson Laroca, Alceu S. Britto Published: 2026-07-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.13345v1 We present an audio-text system for the Ambivalence/Hesitancy Video Recognition Challenge of the 11th ABAW Competition. The method excludes visual frames and represents each video as overlapping 5-second windows aligned with transcript timestamps. Each window combines a 320-dimensional prosodic audio descriptor, a 768-dimensional emotion-oriented RoBERTa embedding, and 74 handcrafted features capturing uncertainty, hedging, and attitudinal conflict. Audio and text are fused via temporal cross-attention, while support features are injected prior to gated multiple-instance learning (MIL) pooling to modulate the window's importance. Predictions from five independently initialized models are averaged. On the labeled public development set, the ensemble achieved an average precision of 0.875 and a macro-F1 of 0.72. Our source code is publicly available at https://github.com/Liga-de-IA-PUCPR/abaw-11-ah-challenge/. --- ### EZSMT Version 3, Matured Authors: Keeran Dhakal, Yuliya Lierler Published: 2026-07-15 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.13344v1 Constraint Answer Set Programming (CASP) is a hybrid reasoning paradigm that combines Answer Set Programming (ASP) with Constraint Processing and Satisfiability Modulo Theories (SMT), enabling powerful declarative encodings of complex combinatorial search problems. This paper presents the design and implementation of EZSMTV3, an extensible SMT-based CASP framework that advances the translational approach to CASP solving. Building upon the foundation of the EZSMT+ system, EZSMTV3 introduces a more expressive input language, supports optimization via weak constraints, and offers foundations for streamlined integration of new constraint types. Rather than implementing custom search procedures, EZSMTV3 leverages state-of-the-art SMT solvers, such as CVC5, YICES, and Z3 to perform reasoning. The paper provides benchmarking results comparing EZSMTV3 with its CASP peers such as CLINGCON, CLINGO(DL), and CLINGO(LP), while showcasing its ability to handle mixed-domain constraints involving both integers and reals. The system provides a robust platform for future extensions and theoretical exploration within the CASP domain. --- ### Marker-free deformable registration and fusion for augmented reality-guided positive margin localization during tumor resection surgery Authors: Yue Yang, Annie Benson, Matthieu Chabanas, Jason Slagle, Thomas Myles, Matthew B. Weinger, Jon S. Heiselman, Michael I. Miga, Michael Topf, Jie Ying Wu Published: 2026-07-15 Categories: cs.CV, cs.ET, cs.HC arXiv: https://arxiv.org/abs/2607.13343v1 Positive margins in head and neck oncologic surgery require mapping specimen-side pathology findings to the patient resection bed. This is challenging because pathologists identify the positive margin on slices of the resected, deformed specimen, while surgeons must relocate the corresponding site on the resection bed using only verbal descriptions and no visual guidance. We present a marker-free augmented reality (AR) workflow for mapping a margin label from a three-dimensional specimen scan to the resection bed. The method combines contour-constrained deformation, residual alignment to a depth scan, surface-based fusion to a head-mounted display, and target projection onto the reconstructed bed. Bead-suture correspondences estimate specimen deformation, whereas patient-to-display fusion does not require external fiducial markers. Following formative experiments, five residents and surgeons performed cadaveric cheek and scalp re-resection tasks under verbal guidance, verbal guidance with specimen examination, and AR guidance. Deformation target errors were $7.63 \pm 3.74$ mm for the cheek and $3.72 \pm 1.02$ mm for the scalp; residual specimen-to-bed distances were $2.43 \pm 2.15$ mm and $2.19 \pm 1.06$ mm, respectively. Fusion error did not differ significantly between marker-free and marker-based methods on either cadaver; overall marker-free fusion error was $2.15 \pm 0.87$ mm. End-to-end margin localization error decreased from $21.40 \pm 3.84$ mm with verbal guidance and $16.09 \pm 4.30$ mm with specimen examination to $6.19 \pm 1.79$ mm with AR guidance ($p < 0.001$). Online fusion required $5.23 \pm 0.34$ s. These results demonstrate effective marker-free AR guidance for positive-margin localization and support more precise tumor resection. --- ### HybridInfer: Thermal-Aware Reinforcement-Learning Tier Routing for On-Device, Edge, and Cloud LLM Inference Authors: Simran Koul Published: 2026-07-14 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.30270v1 On-device inference with small language models keeps user data local, works offline, and incurs no per-query cost, so the on-device tier is preferred when it is adequate. It is thermally constrained, however, and I find the constraint is sharper than a slowdown: on a flagship Snapdragon device, sustained on-device generation destabilizes the GPU inference runtime, which crashes or silently wedges after a few consecutive queries. The failure lies in the current toolchain (OpenCL kernel compilation and long-prompt prefill on the mobile GPU), recurs even when the device is cool, and is worst for long generations. Multi-tier routers across on-device, edge, and cloud models can relieve this pressure, but existing routers are thermal-blind and typically evaluated in simulation or on non-mobile hardware. I present HybridInfer, a thermal-aware reinforcement-learning router for a three-tier hierarchy (on-device Llama 3.2 3B, edge Llama 3.1 8B with retrieval, cloud GPT-4o) that uses the phone's thermal headroom and a query-complexity estimate as state and selects a tier by an offline-trained Q-learning policy. Its reward trades quality against latency, cost, and a thermal penalty, plus a locality bonus crediting on-device execution. I show this bonus is a precondition for thermal-aware routing: without it the optimal policy offloads every query. On a real Android benchmark of 210 prompts, the learned router attains significantly higher quality than two hand-tuned heuristics (paired Wilcoxon, p < 0.02) at the lowest cost of any adaptive condition. Always-on-device conditions match per-query quality on servable queries but are three to six times slower and fail on long queries, so routing wins on latency, reliability, and coverage rather than quality. To my knowledge this is the first use of on-device thermal headroom to select among LLM inference tiers of differing capability on real hardware. --- ### Register Bias in Complexity-Based Large Language Model Routing Authors: Simran Koul Published: 2026-07-14 Categories: cs.CL, cs.CY arXiv: https://arxiv.org/abs/2609.17542v1 Large language model services increasingly route each query to one of several models of differing capability, using a cheap estimate of query complexity to send easy queries to small models and hard queries to large ones. I show that this routing step is not register neutral: text written in a non-standard English register, African American English or the English of second-language writers, is systematically assigned a lower-capacity tier than a meaning-equivalent standard-English version of the same query. The effect is driven by a specific, common routing signal, input length, because non-standard registers omit function words and thus look shorter and therefore simpler; other complexity signals do not carry it. I demonstrate the disparity on 37,704 authentic learner sentence pairs and on a controlled parallel corpus. I then measure the quality consequence on a device, edge, and cloud model ladder and find that the harm is driven by pervasive model bias, every tier, including a frontier cloud model, answers non-standard-register queries significantly less accurately, while the marginal quality cost of the routing decision itself is not significant on this benchmark. Complexity-based routing thus compounds the exposure of the users that the models already serve worst. --- ### MudawanSn: A Gold-Standard Wolof-Arabic Parallel Corpus for Machine Translation Authors: Mouhamed Mbaye, Thierno Diop Published: 2026-07-14 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.17539v1 We present MudawanSn, a gold-standard resource of 1,271 sentence-aligned pairs manually translated from Wolof into Modern Standard Arabic (MSA). The source texts are drawn from the MasakhaNER corpus and cover politics, society, religion, and sports in Senegalese news discourse. Although multilingual resources such as FLORES-200 and NTREX include both Wolof and Arabic, no publicly available parallel corpus is specifically designed for the Wolof-Modern Standard Arabic language pair. We describe the corpus construction protocol, sentence alignment procedure, and quality-control workflow. We benchmark four machine translation systems spanning three architectural families: NLLB-200 (600M), mT5-base, and two AfriNLLB variants, showing that fine-tuning on MudawanSn yields substantial improvements in both translation directions. The best-performing model, AfriNLLB-12, achieves 7.76 BLEU and 30.72 chrF++ for Wolof-to-Arabic, and 8.75 BLEU and 33.08 chrF++ for Arabic-to-Wolof. The corpus is released under the CC BY-NC license and is publicly available on Hugging Face and GitHub. --- ### From Pixels to Pairs: A Comprehensive Benchmark of LLM-Driven Key-Value Extraction in Noisy Document Settings Authors: Zahra Anvari Published: 2026-07-14 Categories: cs.CL, cs.CV arXiv: https://arxiv.org/abs/2609.17538v2 Large language models (LLMs) have demonstrated strong capabilities in document key-value pair (KVP) extraction, yet controlled evaluations of their robustness to optical character recognition (OCR) output remain limited. This leaves an important gap in understanding their reliability in real-world OCR-to-LLM pipelines. Unlike end-to-end Vision-Language Models (VLMs), which jointly perform visual perception and semantic extraction, modular pipelines allow these stages and their errors to be isolated and audited. We introduce a controlled benchmark that distinguishes downstream LLM extraction behavior from upstream OCR degradation. It evaluates 136 experimental configurations and 17,688 document-level inferences generated with deterministic decoding across five instruction-tuned open-weight LLMs (2B-8B parameters), three datasets, and four text-quality conditions. The evaluation combines a full zero-shot comparison, targeted one- to three-shot experiments, and a sensitivity analysis of 40 configurations across 20 frozen demonstration sets. By separating Key Recall (annotated-field recovery) from Exact Match and Value F1 (exact and partial value recovery, respectively), we test whether OCR degradation affects field identification and value reproduction differently across models. Our findings challenge three practical assumptions: (1) clean-text performance reliably predicts real-world robustness, (2) model rankings remain consistent across annotation-derived Gold and OCR-derived text, and (3) additional few-shot demonstrations monotonically improve extraction accuracy. The observed model-ranking reversals and unstable few-shot gains expose important reliability risks under noisy document conditions. We release the benchmarking framework, dataset splits, and evaluation scripts to support reproducible research. --- ### Risk-Aware Decision-Making for Autonomous Overtaking: A World Model-Based Mixture-of-Experts Framework Authors: Yongzhi Liu, Sunan Zhang, Jinchang Xu, Jiawei Wang, Yushu Qiu, Chen Lv, Weichao Zhuang Published: 2026-07-14 Categories: cs.RO, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.00385v1 Autonomous highway overtaking demands foresighted decision-making to handle complex interactions, stochastic traffic evolution, and temporal risk accumulation. However, standard safe reinforcement learning approaches typically rely on implicit value-based risk estimations rather than explicit dynamics modeling, thereby struggling to accurately capture complex risk propagation over multi-step horizons. This limitation frequently results in behaviors that are locally safe but induce substantial latent risks in the long term. To address this, a World Model-based Risk-aware Mixture-of-Experts (WM-RMoE) framework is proposed. First, a learned latent dynamics model facilitates parallel multi-step rollouts, elevating safety assessment from the action level to the trajectory level via cumulative risk evaluation. Second, to enhance robustness under varying interaction intensities, a hierarchical gating mechanism dynamically coordinates experts across long-horizon, short-horizon, and rule-based safety modules. Furthermore, a Gaussian Mixture Model is integrated to preserve multimodal maneuvering branches, thereby mitigating the issue of behavioral mode averaging. Experimental results demonstrate that WM-RMoE significantly outperforms representative baselines in terms of safety compliance, decision stability, and generalization capability. Furthermore, benefiting from the risk-aware formulation, the proposed framework uniquely exhibits the ability to generate foresighted and semantically distinct overtaking maneuvers across diverse traffic densities. --- ### Rating the Raters: Rasch Measurement Theory for LLM Evaluation Authors: Pratik S. Sachdeva, Nathan Boudol Published: 2026-07-14 Categories: cs.AI arXiv: https://arxiv.org/abs/2608.27463v2 LLMs now sit on every side of evaluation: as examinees scored on benchmarks, judges of other models' outputs, and raters of human-generated content. Each paradigm can be viewed as a measurement problem, where a latent property of an object is probed with items from an instrument (e.g., benchmark) by judges or raters. Standard evaluation practices often neglect the contributions of each core component to the end result, limiting our understanding of what is being measured. Rasch measurement theory (RMT) is well-suited to this problem. RMT decomposes ordinal ratings into separable facets on a common scale. It further provides a battery of diagnostics that can identify miscalibrated measurements and rater biases. We present a case study of RMT applied to the LLM-as-rater paradigm using the Measuring Hate Speech corpus, whose construct was itself built under RMT. We fit a series of many-facet Rasch models to annotations from nine LLMs spanning families and capability levels. Our analyses show that LLMs systematically differ from human raters in severity, item-level calibration, question-order robustness, target-identity sensitivity, and rating scale use, all of which standard evaluation practice would largely obscure. Overall, we argue that RMT belongs in the toolkit for evaluating LLM-as-examinee, -judge, and -rater paradigms. --- ### Beyond the Chatbot: Co-Learning and Co-Teaching through a Dual-Persona Generative-AI Assistant Authors: Chaido Mizeli, Marina Delianidi, Konstantinos Diamantaras Published: 2026-07-14 Categories: cs.HC, cs.AI arXiv: https://arxiv.org/abs/2608.24902v1 In this paper we present a generative AI application developed to support both teachers and students in secondary education. The system employs two Large Language Models-LLMs, Gemini and DeepSeek, and a Small Language Model-SLM, Gemma, integrated within a Retrieval Augmented Generation - RAG framework, creating a pedagogically grounded, Greek-language assistant capable of adapting its reasoning and communication style to the user role. Unlike conventional chatbots, the assistant introduces pedagogical persona switching, a dual-role mechanism that enables the same AI model to act as both a teaching companion and a learning guide. Utilizing a RAG paradigm tailored to the Greek educational domain, the architecture segments official textbooks into coherent units. Enriched with specific metadata, these units preserve curricular structure and instructional context, demonstrating how generative AI optimizes modern instructional design. The initial case study focuses on home economics in Greek lower secondary education, a cross-disciplinary subject that integrates elements of economics, health education, and social responsibility. The assistant has been developed to support both learners and educators in complementary ways. In future classroom implementations, students will be able to use it to clarify key concepts such as financial literacy, resource management, and healthy living, while teachers could employ it to design authentic instructional materials, formative assessments, and classroom activities aligned with the official curriculum. The study elevates the concept beyond a simple chatbot, proposing a structured, contextually adaptive framework for pedagogical generative assistants that effectively bridge technology, curriculum, and human learning. --- ### Detection != Reliable Control: Decodable Empathy Directions Yield at Most Partial Shifts in Automated Empathy Scores Authors: Haoran Jisun Published: 2026-07-14 Categories: cs.CL, cs.HC, cs.LG arXiv: https://arxiv.org/abs/2608.24901v1 A decodable "empathy" direction is routinely read as a causal lever, conflating decodability, automated-metric control, and human-perceived change. We test this for two EPITOME-derived facets -- Recognition (cognitive) and Resonance (affective) -- in three instruction-tuned LLMs, scoring every intervention with two LLM judges and a discriminative EPITOME classifier, each gated by an emotional-vs-neutral positive control. The control passes for the affective facet across all automated instruments, but cognitive range is inconsistent across them. Both facets remain decodable after residualizing against a sentence-embedding-derived surface score, and steering can substantially rewrite the text. Yet adding the Resonance direction raises the affective score only partially -- in Qwen by +0.29 (approximately 26% of the natural gap). A direct between-direction contrast confirms the shift is facet-specific in Qwen and Llama (not Gemma); we do not, however, establish a matching human-perceived change. Additive cognitive steering produces no measurable change, but a within-domain control shows the cognitive instrument is too coarse to resolve the differences such steering would produce -- unmeasurable, not a clean null. By contrast, Gemma Recognition ablation lowers the classifier's cognitive score even after adjusting for response length. Detection does not imply reliable control under global interventions, and cognitive-empathy claims warrant an explicit measurement-sensitivity check. --- ### Tradable Itô Signatures: A Model-Free, Interpretable Framework for Dynamic Hedging Authors: Xin Guo, Binnan Wang, Ruixun Zhang Published: 2026-07-14 Categories: q-fin.CP, q-fin.MF, stat.ML arXiv: https://arxiv.org/abs/2608.18120v1 We propose an interpretable machine-learning framework for dynamic hedging using the Itô signature transform, which turns asset-price paths into a set of linear features that universally represent nonlinear functions on time-series. We show that each discretized Itô signature component can be perfectly replicated by a simple self-financing strategy using only the underlying assets and cash, which turns Itô signature components into tradable and transparent hedging bases. This allows nonlinear derivative payoffs to be approximated by linear combinations of signature terms and hedged through the corresponding combination of trading strategies. We further establish a new approximation result for the Itô signature and derive theoretical bounds for both in-sample and out-of-sample hedging errors. Our method is computationally efficient, easy to implement, and avoids the estimation of future conditional expectations, which makes it attractive for real-world applications. In simulations, our method delivers strong sample efficiency at substantially lower computational cost than neural-network benchmarks. In an empirical study of S\&P 500 index options, it performs robustly across vanilla and path-dependent contracts, with the signature-kernel weighted version providing further gains by localizing estimation to similar historical market paths. Overall, the paper identifies the Itô signature as a practical, transparent, and model-agnostic implementation framework for dynamic hedging. --- ### A Human-Centred Approach to Benchmarking LLMs for Parenting Advice Authors: Yunke Zhao, Isobel Voysey, Alastair van Heerden, Rob Hughes, Jun Zhao Published: 2026-07-14 Categories: cs.AI arXiv: https://arxiv.org/abs/2608.14622v1 People are increasingly using large language models (LLMs) to seek advice, including for parenting. Parenting is a critical and socially sensitive domain. Thus, evaluating advice provided by LLMs requires indicators beyond aggregated information quality benchmarks to consider relational and behavioural elements of the responses. With a multi-dimensional rubric created by parenting experts, this paper evaluates 15 LLMs across 100 parenting scenarios in 2 languages (English and Chinese), using an LLM-as-a-judge method. Results show that aggregate scores can hide rubric item-specific weaknesses, models implicitly encourage different parenting styles, and language influences responses. We highlight the importance of evaluation output auditability and challenges involved in evaluating LLM-generated advice in domains like parenting. Our findings provide important insights for selecting LLMs for direct user engagement and the development of user-facing parenting advice applications. --- ### AutoMem: A Text-Gradient Recursive Self-Improvement Framework for Automated Memory Architectures Search Authors: Lin Du, Jie Zhou, Yuxuan Cai, Kai Chen, Qin Chen, Xin Li, Bo Zhang, Wei Li, Liang He Published: 2026-07-14 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2608.14621v1 Long-term memory is increasingly central to LLM agents, yet memory design remains a highly coupled architecture problem: what to encode, how to store it, how to retrieve it, and how to manage it can vary substantially across tasks and backbone models. We construct a discrete search space with 5 encoders, 5 stores, 6 retrievers, and 4 managers, and show that no single memory architecture consistently dominates: different tasks favor different module combinations, leading to substantial performance gaps. Motivated by this, we propose \textsc{AutoMem}, a text-gradient recursive self-improvement framework for task-adaptive memory architecture search. \textsc{AutoMem} optimizes over the factored space through two components: Experience-Guided Architecture Search, which proposes candidate architectures from historical search trajectories and accumulated reflections, and Failure-Guided Module Diagnosis, which localizes memory-related failures to specific modules and converts them into targeted textual feedback. Experiments on GAIA, WebWalkerQA, and xBench-DeepSearch across two LLM backbones show that \textsc{AutoMem} consistently discovers task-adaptive memory architectures that outperform the strongest human-designed memory baselines, improving accuracy by $2.8$ points on average across six benchmark-backbone settings. Further analysis shows that \textsc{AutoMem} achieves a favorable accuracy-efficiency trade-off, reducing token cost by $14.3\%$ over the strongest accuracy baselines under Qwen3.5-122B-A10B, while also finding stronger architectures than substantially larger random searches within only a few guided iterations. --- ### Robust XGBoosting for Regression Authors: Iris Aragón Mladosich, Christophe Croux Published: 2026-07-14 Categories: cs.LG, stat.CO, stat.ML arXiv: https://arxiv.org/abs/2608.13590v1 XGBoost is a very popular and powerful method for prediction. It iteratively fits simple decision trees to the residuals of the previous step. An efficient and scalable implementation is available. The standard loss function for XGBoost is the quadratic loss, but a Huber loss can also be used. In this paper, we study the robustness of XGBoost and show that its performance can be affected by vertical outliers and leverage points. To address this, we explore alternative loss functions, based on M-, S-, and τ -estimators from robust regression. Our results indicate that a two-step procedure, referred to as MM-XGBoost, provides the best trade-off between robustness and prediction accuracy. --- ### Agreement Is Not Alignment: Divergent Moral Grounds in Human and LLM Ethical Judgments Authors: Octavian M. Machidon, Alina L. Machidon, Vojko Strahovnik, Mateja Centa Strahovnik, Jonas Miklavčič, Marko Robnik Šikonja Published: 2026-07-14 Categories: cs.AI arXiv: https://arxiv.org/abs/2608.12368v1 Agreement with human judgments is a common proxy for evaluating the alignment of large language models (LLMs). Yet agreement in final labels does not show that human annotators and models rely on the same moral grounds. Two agents may reach the same judgment while appealing to different principles, contextual assumptions, or interpretations of the situation. We test this distinction using a curated 500-item ETHICS-derived benchmark spanning five domains of moral judgment, with new human annotator and LLM annotations of both final labels and supporting rationales. Across frontier and open model families, agreement with human annotator majority labels is often high. However, rationale-level analysis reveals systematic divergence in the moral grounds expressed by human annotators and models. In particular, models redistribute attention across categories such as harm, respect, promise-keeping, justice, desert, and excuse relevance, even when their final labels match the human annotator majority. Our results show that agreement should not be treated as equivalent to alignment. Label-based evaluation can therefore be misleadingly reassuring unless complemented by analysis of the reasons, principles, and moral priorities expressed in model judgments. --- ### Detecting a Route Flip Is Easier Than Knowing Whether to Fix It: Causal Route-Mediated Damage in Quantized Mixture-of-Experts Authors: Parvel Gu Published: 2026-07-14 Categories: cs.AI, cs.CL, cs.LG arXiv: https://arxiv.org/abs/2608.11212v1 Top-k Mixture-of-Experts (MoE) routing is discontinuous, so a deployment-motivated numerical disturbance -- simulated 4-bit KV-cache quantization read by a protected BF16 gate -- pushes tokens across decision boundaries and flips which experts fire. This paper proposes no new mitigation; it supplies a causal apparatus, empirical findings, and a detection-limit result. A four-run apparatus prices the route-mediated fraction (RMF) of quantization damage, a token-level attribution decomposes it by mechanism, and pre-registered probes carry the findings across three architectures. On OLMoE-1B-7B at 4-bit KV (pilot), about a third of the damage is routing-mediated: RMF ~ 0.31 (discovery 0.31 (0.20, 0.41); process-replicated mean 0.313 +/- 0.020; pre-registered re-execution 0.231). The deployable router margin detects that a flip occurred (AUC 0.772) but cannot tell a harmful flip from a helpful one (at chance): among the tested local, inference-observable router statistics we find no predictor of a flip's loss sign above chance -- an empirical benefit-detection barrier bounding selective repair restricted to this feature family. The signed-flip tax and sign-inseparability carry cross-model; the clean-reference remedy's payout is architecture-modulated; a controlled same-checkpoint flag-swap re-scopes the gate's normalization convention to a damage-magnitude moderator, not a route-recoverability mechanism. A real int4 KV kernel yields a fraction compatible with the fake-quant dose curve but underpowered (95% CI (-0.111, 0.394) includes zero) -- ruling out gross disagreement, not an independent replication. Hypotheses, thresholds, and evaluations were pre-registered before measurement, with misses reported; a pre-registered held-out read replicates the partition and the near-cancelling tax out of sample, while the strict impossibility exclusion narrowly misses. --- ### Training Variable Long Sequences with Data-Centric Parallel Authors: Geng Zhang, Xuanlei Zhao, Kai Wang, Yang You Published: 2026-07-14 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2608.07524v1 Training deep learning models on variable long sequences poses significant computational challenges. Existing methods force a difficult trade-off between efficiency and ease-of-use. Simple approaches use static configurations that cause workload imbalance low efficiency, while complex methods introduces significant complexity and code change for new models. To break this trade-off, we introduce Data-Centric Parallel (DCP). Its core principle is to let the data itself drive the runtime. It achieves this by dynamically adjusting direct runtime settings (e.g., parallel size, gradient accumulation, recomputation) based on each batch's sequence length. Empirical results demonstrate that our method achieves up to a 2.88$\times$ speedup on 32 H200 GPUs. Designed for generalization, it can be integrated into any model with 10 lines of code. We anticipate this simple yet effective approach will serve as a robust baseline and facilitate future advancements in distributed training for variable long sequences. --- ### Width, Memory, and Delay: A Resource Accounting for the Limits of Flat Multi-Agent Systems Authors: Oleksandr Kuznetsov, Emanuele Frontoni Published: 2026-07-14 Categories: cs.MA, cs.AI arXiv: https://arxiv.org/abs/2608.00028v1 A recurring question in the design of scalable multi-agent systems -- from robot swarms to collectives of large-language-model (LLM) agents -- is whether adding more agents can, on its own, overcome performance limits, or whether a qualitatively \emph{deeper} organization is required. A recent preprint argues that flat, homogeneous multi-agent systems face an irreducible, population-independent ``causal floor'' on achievable error, removable only by hierarchical (nested-loop) organization. Using a controlled disturbance-rejection testbed with an exactly computable optimum, we show this conclusion is too strong and replace it with a quantitative resource model built on three resources: population \emph{width} $N$, per-agent internal-model \emph{memory} $d$, and prediction across the observation \emph{delay} $τ$. We establish three claims. (i) The achievable floor is governed not by architectural hierarchy but by per-agent internal-model content: a flat, homogeneous swarm whose agents carry a matched internal model of the disturbance matches or beats a designed two-loop hierarchy at equal per-agent memory -- so temporal depth can be dynamical (recurrent memory), not architectural (nesting). (ii) The three resources are \emph{not mutually interchangeable}; we chart the exchange rates and the hard non-exchange boundaries on an explicit width$\times$memory map, including a strict equal-total-state-budget comparison. (iii) A residual floor is set by the observation delay and the environment's unpredictability over that horizon, which we verify against the optimal controller. We quantify the price of replacing oracle knowledge of the disturbance spectrum with online learning, provide a preliminary robustness check against a mild bounded nonlinearity and a spatially-extended plant, and distill four design rules for practitioners. --- ### Motif-Mamba: network motif improved mamba for long-range sequence modeling Authors: Chonghe Hao, Yue Sun, Jian Zhang, Yansong Wang, Wangzi Yao, Yunjie Yao, Tielin Zhang Published: 2026-07-14 Categories: cs.AI arXiv: https://arxiv.org/abs/2608.00027v1 Efficient long-sequence modeling remains a central challenge for large language models, as self-attention scales quadratically with sequence length. Mamba offers a linear-time alternative through selective state space recurrence, but its predominantly diagonal state transitions restrict explicit interactions among state dimensions. We propose Motif-Mamba, a structured state space model that augments Mamba with a motif-constrained low-rank recurrent pathway. Inspired by the dynamics of three-node network motifs, the proposed pathway projects hidden states into a compact dynamical subspace, imposes motif-guided interactions, and maps the resulting dynamics back to the original state space. This design enhances cross-dimensional communication while preserving the linear-time recurrent structure of Mamba. Experiments on long-sequence extrapolation, language modeling benchmarks, and brain--computer interface decoding show consistent improvements over Mamba backbones, suggesting that motif-guided low-rank dynamics provide an effective structural prior for long-range sequence modeling. --- ### AI-assisted pre-review of open-source software submissions: an experience report from BOSC 2026 Authors: Tazro Ohta, Nomi L. Harris, Seth Carbon Published: 2026-07-14 Categories: cs.CL, cs.AI, cs.DL arXiv: https://arxiv.org/abs/2607.27228v1 Most conferences rely on peer-review of submissions, but as generative AI makes it easier than ever to prepare submission materials, some conferences are seeing an overwhelming surge of submissions. We wanted to see if generative AI could help our conference's volunteer reviewers by pre-reviewing abstracts for certain criteria. The Bioinformatics Open Source Conference (BOSC) was well-positioned to experiment with this, as we already had a detailed rubric used by reviewers to evaluate submitted abstracts on multiple criteria, including openness (public availability of the code or other content associated with the project), valid open source license, and "runnability" (how easy it is to download, build, and run the project - an important measure of reusability). For BOSC 2026, we built bosc-pre-review, an agentic skill that assessed six review criteria, and Runabilly, which builds and tests each project in a disposable Docker container for safety. The AI only gathered evidence to present to the reviewers; humans made every decision regarding the acceptance of the abstracts. After the review period, we surveyed the reviewers to determine how useful they found the pre-review. Most of those who responded said they found it useful, but they preferred to check the AI's conclusions against their own, rather than accepting the AI results unquestioningly. --- ### FinCacheServe: Dependency-Consistent Answer Reuse for Cost-Efficient RAG Serving over Mutable Enterprise Documents Authors: Lingteng Zeng, Yifan Jin Published: 2026-07-14 Categories: cs.IR, cs.AI arXiv: https://arxiv.org/abs/2607.26076v1 Retrieval-augmented generation services over mutable enterprise documents repeatedly execute semantically equivalent analysis requests. Answer reuse can remove GPU-bound generation work, yet response caches require dependency consistency when filings, evidence chunks, and tool outputs change. FinCacheServe treats each generated answer as a serving object indexed by enterprise intent and guarded by document versions, evidence fingerprints, tool fingerprints, model identity, and decoding configuration. A vLLM implementation evaluates SEC-derived financial-document workloads with Qwen2.5 models. On a 2,230-request hosted 7B trace, FinCacheServe skips 53.27% of LLM calls with zero observed dependency-stale outputs. Across three hosted 32B operator-suite seeds, it skips 53.31% of 544 requests, compared with 38.97% for versioned semantic caching and 22.43% for grounded-style reuse. Capacity, backend, and SLO replays show oracle-bounded cache management, 100k-entry transactional metadata behavior, and 44.30% lower estimated Wh per dependency-fresh 2s-SLO success than versioned semantic caching. --- ### Aletheia: An Offline-First Clinical Decision Support System for Differential Diagnosis in Low-Resource Healthcare Settings Authors: Joseph Walusimbi, Ann Move Oguti, Abubakhari Sserwadda, Precious Boss Kasasira, Charles Brian Okoboi Published: 2026-07-14 Categories: cs.AI, cs.LG, q-bio.OT arXiv: https://arxiv.org/abs/2607.24814v3 Access to specialist clinical expertise remains severely limited across sub-Saharan Africa, where physician-to-patient ratios can fall below 1:25,000 in rural settings. Existing AI-assisted diagnostic tools predominantly require reliable internet connectivity and high-specification hardware, rendering them impractical for frontline healthcare workers in district hospitals and health centres. This paper presents Aletheia, an offline-first clinical decision support system designed for low-resource healthcare contexts across sub-Saharan Africa. Aletheia is built upon Qwen2.5-3B-Instruct, fine-tuned using Quantised Low-Rank Adaptation (QLoRA) on a curated dataset of 27,000 clinical reasoning samples spanning 50 disease conditions with elevated prevalence in East Africa. Evaluation demonstrates a Top-1 diagnostic accuracy of 80% (8 of 10 cases; 95% CI: 49.0-94.3%), Top-3 accuracy of 100% (10 of 10 cases; 95% CI: 72.2-100%), BERTScore-F1 of 0.909, and METEOR of 0.467. These diagnostic figures are computed over a deliberately small set of ten representative clinical case categories, one case each, and are therefore indicative rather than statistically robust; the wide confidence intervals should be read alongside them. The system achieves an Expected Calibration Error (ECE) of 0.275 and passes the Africa Deep Tech Challenge 2026 (ADTC 2026) memory budget constraint of 7168 MB, achieving a peak inference RAM of approximately 3630 MB on the standardised benchmark laptop. These results demonstrate the feasibility of deploying large language model-based clinical reasoning at the primary care level in resource-constrained settings without cloud infrastructure. --- ### Aletheia: An Offline-First Clinical Decision Support System for Differential Diagnosis in Low-Resource Healthcare Settings Authors: Joseph Walusimbi, Ann Move Oguti, Abubakhari Sserwadda, Precious Boss Kasasira, Charles Brian Okoboi Published: 2026-07-14 Categories: cs.AI, cs.LG, q-bio.OT arXiv: https://arxiv.org/abs/2607.24814v1 Access to specialist clinical expertise remains severely limited across sub-Saharan Africa, where physician-to-patient ratios can fall below 1:25,000 in rural settings. Existing AI-assisted diagnostic tools predominantly require reliable internet connectivity and high-specification hardware, rendering them impractical for frontline healthcare workers in district hospitals and health centres. This paper presents Aletheia, an offline-first clinical decision support system designed for low-resource healthcare contexts across sub-Saharan Africa. Aletheia is built upon Qwen2.5-3B-Instruct, fine-tuned using Quantised Low- Rank Adaptation (QLoRA) on a curated dataset of 27,000 clinical reasoning samples spanning 50 disease conditions with elevated prevalence in East Africa. Evaluation demonstrates a Top-1 diagnostic accuracy of 80.0%, Top-3 accuracy of 100.0%, BERTScore-F1 of 0.909, and METEOR of 0.467 across ten representative clinical case categories. The system achieves an Expected Calibration Error (ECE) of 0.275 and passes the Africa Deep Tech Challenge 2026 (ADTC 2026) memory budget constraint of 7 168 MB, achieving a peak inference RAM of approximately 3 630 MB on the standardised benchmark laptop. These results demonstrate the feasibility of deploying large language model-based clinical reasoning at the primary care level in resource-constrained settings without cloud infrastructure. --- ### Towards Nexus-Score: Metadata Gaps Limit Scholarly AI Attribution Authors: Aadi Narayana Varma Dantuluri, Sushrut Thorat, Paras Chopra Published: 2026-07-14 Categories: cs.DL, cs.AI, cs.IR arXiv: https://arxiv.org/abs/2607.22684v1 Artificial intelligence systems increasingly mediate how science is found and credited. We asked whether missing metadata prevents AI systems from crediting work. As a boundary test, an AI system citing without access to task-relevant paper lists often produced out-of-list identifiers, some fabricated. We then tested the mechanism in real scholarly infrastructure by using OpenAlex records to hide or restore author, institution, funder, reference, and text-access links while holding works and tasks fixed. Restoring the relevant link made the corresponding attribution possible; restoring the wrong kind did not, with 0 correct answers across 469 completed mismatched tests. Thus, in these tasks, one metadata facet did not substitute for another. Missing links led to invented answers, refusals, or tool-budget exhaustion, and web search did not recover hidden author links. In sum, AI systems credited work only when record connections were visible or recoverable. This motivates Nexus-Score, a record-level check for metadata gaps, to guide repair and help prepare the scholarly record for AI-mediated use. --- ### Beyond SBDD: Geometric Deep Learning in Polypharmacology and Multi-target Drug Design Authors: Tianming Han, Zhijie Pan, Wenchi Ge, Qi Zhao Published: 2026-07-14 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2607.20550v1 The traditional "one drug, one target" paradigm of structure-based drug design (SBDD) frequently proves inadequate for treating multifactorial diseases such as cancer and neurodegenerative disorders, owing to compensatory signaling pathways and the emergence of drug resistance. While polypharmacology offers a synergistic therapeutic strategy, the rational design of ligands capable of simultaneously satisfying the geometric constraints imposed by multiple targets remains a major computational bottleneck. This review positions geometric deep learning (GDL) as a powerful integrative approach to overcome these limitations. We systematically survey GDL architectures ranging from invariant graph neural networks to SE(3)-equivariant diffusion models that harness non-Euclidean molecular data to capture intrinsic three-dimensional (3D) structural interdependencies. We critically analyze GDL applications across three core dimensions, including the characterization of shared binding pockets via geometric embeddings, multi-target bioactivity prediction through heterogeneous graph fusion, and de novo generation of dual-target ligands. Particular emphasis is placed on emerging structure-conditioned generative algorithms that integrate diffusion models with reinforcement learning to autonomously resolve complex geometric conflicts between competing binding sites. Furthermore, we evaluate the pivotal role of multimodal omics integration and specialized geometric benchmarking infrastructures in validating these models. By synthesizing these methodological advances, this review elucidates the paradigm shift in drug discovery from serendipitous exploration to rational, structure-driven polypharmacological molecular engineering, thereby providing a clear, structured guide for navigating the complexities of next-generation therapeutics. --- ### On Incentivized Exploration beyond Bayesianism and Full-Information Authors: Dimitar Chakarov, Lee Cohen, Nathan Srebro Published: 2026-07-14 Categories: cs.GT, cs.LG arXiv: https://arxiv.org/abs/2607.18300v1 We extend Incentive Compatible Exploration beyond the Bayesian full-information setting of Kremer et al. (2014). We consider agents that may possess external information unknown to the principal. We show such settings require new notions of incentivized exploration, as well as going beyond a Bayesian perspective, and we introduce a definition where agents choose any reasonable (undominated) action. Furthermore, our framework provides for a more robust treatment of ties, and extends to settings where agents lack a single common prior and instead only know that reward distributions belong to a collection of potential priors. --- ### Generalist AI Control: Towards Multi-purpose Adaptive Algorithms Authors: Klinsmann Agyei, Pouria Sarhadi Published: 2026-07-14 Categories: cs.AI, cs.MA, cs.RO, eess.SY arXiv: https://arxiv.org/abs/2607.16313v1 Traditional controllers are designed for specific systems and do not transfer across different system orders and dynamics. We present a Generalist Controller, a learning-based controller capable of controlling systems of varying orders and dynamics. The approach introduces a novel dynamic state-space representation using attention mechanisms with masking, enabling a single neural network, trained in one shot, to handle systems with different dimensions without architectural modifications by assigning a system tag to each system. We generated 314,630 demonstrations from 25 diverse systems, including stable, unstable, minimum-phase, and non-minimum-phase dynamics, spanning linear and nonlinear systems from autonomous underwater and aerospace vehicles to mechanical systems and chemical processes. The model learns cross-system control strategies through multi-scale temporal processing and a mixture-of-experts architecture. Simulation results demonstrate that the proposed generalist controller achieves comparable performance to system-specific LQI controllers across all tested systems, including challenging cases such as non-minimum-phase and unstable dynamics, whilst generalising to unseen operating conditions including actuator saturation, noise, disturbance, and reference trajectories not encountered during training. This work represents a significant step towards generalist control policies within a defined family of dynamical systems, demonstrating effective control across a range of single-input single-output (SISO) systems of varying order and dynamics using a single learned policy without system-specific tuning. --- ### xperception -- Making Robotic Grasping Easier Authors: Matteo Bortolon, Andrea Caraffa, Alice Fasoli, Fabio Poiesi Published: 2026-07-14 Categories: cs.CV, cs.RO arXiv: https://arxiv.org/abs/2607.16312v1 The transition toward high-mix low-volume manufacturing demands flexibility in robotic manipulation. However, conventional vision systems remain a bottleneck, requiring extensive data collection and model retraining whenever a new object is introduced to the production line. To overcome this rigidity, we present xperception, a zero-shot 6D pose estimation technology that eliminates the need for object-specific fine-tuning and laborious data annotation. By directly utilizing typical CAD models and integrating the rich semantic features of foundation models (e.g. DINOv2, GeDi), xperception achieves millimeter-accurate 6D pose estimation. xperception showed robustness against severe occlusions in industrial tasks like bin picking and is engineered for deployment on industrial edge hardware, such as NVIDIA Jetson Thor. Validated at a TRL of 6, the core methodology behind xperception is based on the FreeZe algorithm, which won the international BOP Challenge 2024, paving the way for scalable, plug-and-play robotic automation in unstructured high-mix low-volume manufacturing industries. --- ### Seeing What Is Actually There: PriVE-Bench and PriVE-Tools for Counterfactual Evaluation of Agentic Visual Evidence in VLMs Authors: Jingyu Sun, Jiachen Tu, Yuyang Xue, Yaoxin Jiang, Guoyi Xu, Zhengtao Yao, Rui Qian, Yizheng Sun, Hongpeng Zhou, Jingyuan Sun, Yan Lin Published: 2026-07-14 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2607.16311v1 Vision-language models (VLMs) often answer visual questions using learned language and category priors rather than grounding their predictions in the image itself. Counterfactual images provide a natural diagnostic setting for this failure mode: when visible evidence contradicts what is usually true, a grounded model should answer from the pixels, while a prior-following model will produce a canonical but visually incorrect response. However, existing counterfactual benchmarks mainly ask whether such prior-following behavior exists. In this paper, we ask a further question motivated by the rise of tool-augmented and agentic vision systems: can additional visual evidence views help VLMs reason against their priors? We introduce PriVE-Bench, a Prior-vs-Visual Evidence Benchmark that uses paired original and counterfactual images to distinguish visually grounded answers from prior-consistent errors. We further introduce PriVE-Tools, a controlled agentic-vision-inspired extension that evaluates whether tool-derived visual evidence -- including bounding boxes, crops, zoom panels, and contours -- improves grounding under the same counterfactual conflicts. Across open- and closed-source VLMs, we compare raw, paired-image, and tool-conditioned inputs using accuracy, prior-following error rate, and other-response rate. Our results show that visual evidence tools can help in some settings, especially when models can use localized evidence effectively, but they are not a universal remedy: several models continue to follow language and category priors even when relevant visual evidence is explicitly provided. --- ### SaaF: Scene-Specific Ambiguity-Aware 3D Language Fields towards Interactive Real-World Object Retrieval Authors: Yuga Yano, Daiju Kanaoka, Hakaru Tamukoh, Yasutomo Kawanishi Published: 2026-07-14 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.16309v1 We propose Scene-specific Ambiguity-aware 3D Language Fields (SaaF), a novel Gaussian Splatting-based 3D language field designed for interactive object retrieval in a given real-world scene. Interactive object retrieval using natural language is a crucial capability for service robots operating in complex real-world environments. While recent 3D language field methods for object retrieval establish associations between rendered pixels and autoencoder-compressed CLIP features, they suffer from two limitations: (1) reduced discriminability among similar objects due to feature compression, and (2) poor handling of ambiguous queries, often resulting in unstable or incorrect retrieval. To address these limitations, SaaF introduces a metric learning strategy to construct a unified feature space that is both instance-discriminative and ambiguity-aware. (i) To enhance instance-level visual discrimination, SaaF employs metric learning that pulls image features from multiple viewpoints of the same object closer together in the feature space. (ii) To establish ambiguity awareness, the model jointly trains on multiple text labels generated by the proposed method from each tracked object image sequence, including ambiguous descriptions, to learn the semantic relationships between ambiguous and specific features in a target scene. This feature space enables fine-grained visual understanding while allowing the system to estimate query ambiguity and interactively request clarification when needed. Experimental results demonstrate that SaaF not only improves retrieval accuracy over previous methods but also robustly detects and handles ambiguity in the user text queries under open-vocabulary settings. --- ### DAUPNet: Domain-Aware Uncertainty Modeling for Reliable Prototype Discrimination in Cross-Domain Few-Shot Semantic Segmentation Authors: Lei Yuan, Zhongxu Hu, Jingyi Wen, Pengxing Yi Published: 2026-07-14 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2607.16308v1 Cross-domain few-shot semantic segmentation (CD-FSS) has predominantly been formulated as learning domain-invariant representations or improving support-query correspondence. Nevertheless, large domain shifts still make prototype matching unreliable: inconsistent hierarchical responses corrupt the support representation, deterministic prototypes cannot express boundary and appearance ambiguity, and treating prototypes with different reliability equally during optimization weakens foreground-background separation. We therefore propose DAUPNet, a unified framework that reformulates cross-domain prototype matching as uncertainty-aware prototype discrimination. DAUPNet first harmonizes hierarchical support-query features to provide stable evidence, then represents foreground and background prototypes probabilistically, and finally uses their estimated uncertainty to regulate contrastive optimization. On four standard target domains, DAUPNet achieves 72.6% and 76.7% average mIoU in the 1-shot and 5-shot settings, respectively, including substantial gains on the two medical domains. These results demonstrate that modeling prototype uncertainty and incorporating it into optimization provides a robust and interpretable approach to CD-FSS under severe domain shift. The code is available at https://github.com/madness-Lei/DAUPNet --- ### CoBind: Stage-Aware Compositional Binding for Training-Free Text-to-Image Generation Authors: Kaijie Chen, Ethan Caldwell, Mira Vossen, Julian Hartwell, Serena Whitlock, Adrian Bellamy Published: 2026-07-14 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.16307v1 Diffusion-based text-to-image models often fail on complex prompts involving multiple entities, attributes, and relations, producing object omissions, incorrect attribute assignments, or reversed spatial layouts. Existing training-free methods mainly strengthen token-level attention, but do not explicitly model which attributes belong to which entities or when different constraints should be enforced during denoising. We introduce \textbf{CoBind}, a training-free framework for stage-aware compositional binding. CoBind parses a prompt into a composition graph of entities, attributes, and relations. It first establishes the global layout using entity-completeness and relation constraints, then binds attributes to their target entities through contrastive cross-entity optimization. Structural guidance is gradually relaxed in later denoising steps to preserve textures and visual details. CoBind also adapts the guidance strength according to the current satisfaction of each constraint, reducing unnecessary latent updates. CoBind requires no retraining or additional annotations. Experiments on T2I-CompBench++, GenEval, and multiple diffusion backbones show consistent improvements in attribute binding, spatial relations, and complex compositional generation while maintaining competitive visual quality. --- ### LookME: Lookup-Based Multimodal Embeddings for Layer Injection in Vision-Language Models Authors: Zeyu Xu, Xingzhong Hou, Pengkai Guo, Siling Lin, Xiao Xu, Menghua Zhai, Haoyu Chen, Yunke Zhang, Fei Huang Published: 2026-07-14 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2607.16305v2 Vision-Language Models (VLMs) have achieved strong progress in multimodal understanding. However, scaling dense or sparse Mixture-of-Experts (MoE) models to improve performance limits deployment in resource-constrained environments due to the trade-off between high memory usage from full loading and increased latency from on-demand loading. Recently, the Per-Layer Embedding (PLE) architecture addresses this by scaling models with large external embedding tables stored in read-only memory (ROM) and performing lightweight lookup to retrieve relevant embeddings to enhance token representations. Nevertheless, existing PLE-style methods are primarily designed for text embeddings due to the convenience of ID-based retrieval, limiting their effectiveness in VLMs where multimodal embeddings contain richer information for visual tasks. In this paper, we propose LookME, the first framework that enables lookup-based enhancement for multimodal embeddings in VLMs while supporting partitioned storage and on-demand loading. To efficiently lookup arbitrary continuous multimodal embeddings from large-scale embedding tables, we propose a hierarchical two-level lookup method employing a coarse-to-fine strategy that performs lookups from the scene-level to the intra-scene primitive-level. Furthermore, we integrate the lookup method with a sparse injection strategy, which adaptively prioritizes critical embeddings over voluminous multimodal embeddings within layers, and facilitates embedding table reuse across neighboring layers, improving the trade-off among efficiency, model size, and performance. Experiments on multiple visual benchmarks show that LookME outperforms text-only PLE-style methods, validating the effectiveness of lookup-based multimodal embedding enhancement. --- ### LookME: Lookup-Based Multimodal Embeddings for Layer Injection in Vision-Language Models Authors: Zeyu Xu, Xingzhong Hou, Pengkai Guo, Siling Lin, Xiao Xu, Menghua Zhai, Haoyu Chen, Yunke Zhang, Fei Huang Published: 2026-07-14 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2607.16305v1 Vision-Language Models (VLMs) have achieved strong progress in multimodal understanding. However, scaling dense or sparse Mixture-of-Experts (MoE) models to improve performance limits deployment in resource-constrained environments due to the trade-off between high memory usage from full loading and increased latency from on-demand loading. Recently, the Per-Layer Embedding (PLE) architecture addresses this by scaling models with large external embedding tables stored in ROM and performing lightweight lookup to retrieve relevant embeddings to enhance token representations. Nevertheless, existing PLE-style methods are primarily designed for text embeddings due to the convenience of ID-based retrieval, limiting their effectiveness in VLMs where multimodal embeddings contain richer information for visual tasks. In this paper, we propose LookME, the first framework that enables lookup-based enhancement for multimodal embeddings in VLMs while supporting partitioned storage and on-demand loading. To efficiently lookup arbitrary continuous multimodal embeddings from large-scale embedding tables, we propose a hierarchical two-level lookup method employing a coarse-to-fine strategy that performs lookups from the scene-level to the intra-scene primitive-level. Furthermore, we integrate the lookup method with a sparse injection strategy, which adaptively prioritizes critical embeddings over voluminous multimodal embeddings within layers, and facilitates embedding table reuse across neighboring layers, improving the trade-off among efficiency, model size, and performance. Experiments on multiple visual benchmarks show that LookME outperforms text-only PLE-style methods, validating the effectiveness of lookup-based multimodal embedding enhancement. --- ### Dual-Domain Self-Supervised Artifact Removal Framework for Photoacoustic Computed Tomography Authors: Yucheng Zhou, Shuang Li, Yu Zhang, Yibing Wang, Chulhong Kim, Seongwook Choi, Changhui Li Published: 2026-07-14 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.16304v1 Photoacoustic Computed Tomography (PACT) often faces severe challenges from reconstruction artifacts due to sparse detection conditions. In this work, based on the distinct differences in artifact patterns between back-projection-based and Fourier-based reconstruction algorithms, we propose a self-supervised artifact removal framework that employs a lightweight Siamese Neural Network and a composite loss function integrating cross-domain fidelity and uncertainty-weighted consistency, effectively decoupling dual-domain features and filtering artifacts. Comprehensive validations using simulations, phantoms, in vivo rat and human experimental data demonstrate that the proposed method can significantly suppress image artifacts. Furthermore, enabled by the acceleration of the spatial-domain and frequency-domain inverse operator, this end-to-end approach also achieves exceptional computational efficiency. --- ### Med-OPD: Improving Medical Vision-Language Models via Evidence-Aware On-Policy Distillation Authors: Yunhang Qian, Jiaquan Yu, Jiawei Liu, Meng Wang, Hongwei Bran Li, Xiaobin Hu Published: 2026-07-14 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2607.16303v1 Medical Vision-Language Models (Med-VLMs) require reliable reasoning from fine-grained visual evidence, yet existing models can produce plausible clinical answers by relying on language priors or medical templates rather than truly attending to diagnosis-critical regions. On-Policy Distillation (OPD) offers dense token-level supervision on student-generated trajectories and provides a privacy-compatible means of capability transfer without requiring the redistribution of raw patient data. However, standard OPD uniformly distills all tokens, causing sparse evidence-dependent tokens to be diluted by abundant clinical narrative tokens. Inspired by the success of OPD in the large language model community, we propose \textbf{Med-OPD}, to our knowledge the first unified post-training framework that integrates on-policy distillation with medical evidence-aware supervision for Med-VLMs. We introduce \textbf{Medical Evidence Advantage} (MEA), a teacher-grounded counterfactual signal that uses an answer-aware hint to focus teacher scoring on evidence supporting the target diagnosis, and measures each token's dependence on medical visual evidence by comparing teacher likelihoods under the original and evidence-degraded imaging modalities. Based on MEA, Med-OPD redistributes the distillation signal at both the token and trajectory levels, emphasizing diagnosis-critical tokens and evidence-reliant rollouts. Experiments on OmniMedVQA subsets show that Med-OPD consistently outperforms SFT and standard OPD across CT, MRI, Disease Diagnosis, and Lesion Grading. These results demonstrate that evidence-aware distillation can better strengthen medical VLMs' reliance on key visual evidence and improve reliable multimodal medical reasoning. The source code and data is publicly available at: https://github.com/yunhang8658/MedOPD.git --- ### Natural-Language to SysMLv2 Translation via Conformance-Driven Iterative Refinement Authors: Chance LaVoie, Eladio Andujar Lugo, Taylan G. Topcu, Levent Burak Kara Published: 2026-07-14 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2607.14162v1 Model-Based Systems Engineering (MBSE) relies on formal system models as primary technical artifacts for representing requirements, structure, and behavior across the system lifecycle. With the standardization of SysMLv2 as a textual language, interest is increasing in translating natural-language descriptions directly into executable models. For practical deployment, generated models must be accepted by industrial modeling environments, not merely satisfy grammar constraints. We present a conformance-checker-driven framework for reliable natural-language-to-SysMLv2 translation that enforces production-level acceptance as the termination condition. The system embeds a SysMLv2 conformance checker within a generate-check-repair loop. Each model is evaluated using the checker, and deterministic diagnostics are incorporated into revisions until zero conformance errors are achieved. Using the production checker as the oracle ensures the framework targets deployability rather than grammar plausibility. We evaluate the approach on the full SysMBench prompt set of 151 prompts across four large language model backends, yielding 604 prompt-model cases. Single-shot generation achieves 51.16% production-conformance acceptance, while our approach achieves 100.00% conformance. By elevating production conformance from a post-processing check to a control mechanism within generation, the framework converts probabilistic outputs into production-accepted SysMLv2 artifacts suitable for loading, visualization, and engineering use. --- ### Deep-learning Causal Retrieval Optimization for Efficient e-commerce Distribution in Pinterest Authors: Junpeng Hou, XianXing Zhang, Sai Xiao, Derek Cheng, Darren Reger, Olafur Gudmundsson, Mehdi Ben Ayed, Zhiqing Rao, Huizhong Duan Published: 2026-07-14 Categories: cs.IR, cs.LG arXiv: https://arxiv.org/abs/2607.14161v2 Pinterest is where people turn inspiration into action as users browse ideas, then take steps toward realization, often by discovering shoppable content. To support this journey, we must distribute commerce content when it helps, not when it distracts. We frame this as a causal decision of triggering shopping candidate generators in early retrieval and deploy a production system at Pinterest that learns personalized and contextualized triggering policies. A deep multi-task model jointly predicts outcomes and uplift of multiple events, trained with a doubly-robust pseudo-outcome alongside calibrated outcome losses for stable, single-robust uplift learning. A randomized data logging supplies counterfactual coverage, and the model is evaluated by both regular and reverse metrics for full assessment. A linear-time offline replay is designed to select thresholds and forecast policy impact with extremely high consistency with online results. For productionization, the model runs in parallel with remote retrieval calls without end-to-end latency regression. At web scale, we cut shopping triggers by up to 85% while holding key shopping sessions neutral, improving important total sessions (+0.26%) and Pin saves (+1.10%), with significant infrastructure savings. By unifying deep causal learning with reliable offline replay and demonstrating production-grade deployment, this work provides a generally practical recipe for early-retrieval optimizations in modern cascading recommenders beyond shopping, aligning exploration and cost with user intent at scale. ---