# Research Papers — 2025-12 · bundle page 30 of 175 > Full-text markdown of papers #1451–1500 of 8703 published in 2025-12, newest first. AI-readable mirrors of the full 8703-paper 2025-12 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2025-12/page/30 [← Previous bundle](https://aipod.dev/papers/archive/2025-12/page/29.md) [Next bundle →](https://aipod.dev/papers/archive/2025-12/page/31.md) ### SpidR-Adapt: A Universal Speech Representation Model for Few-Shot Adaptation Authors: Mahi Luthra, Jiayi Shen, Maxime Poli, Angelo Ortiz, Yosuke Higuchi, Youssef Benchekroun, Martin Gleize, Charles-Eric Saint-James, Dongyan Lin, Phillip Rust, Angel Villar, Surya Parimi et al. Published: 2025-12-24 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2512.21204v2 Human infants, with only a few hundred hours of speech exposure, acquire basic units of new languages, highlighting a striking efficiency gap compared to the data-hungry self-supervised speech models. To address this gap, this paper introduces SpidR-Adapt for rapid adaptation of speech units to new languages using minimal unlabeled data. We cast such low-resource speech representation learning as a meta-learning problem and construct a multi-task adaptive pre-training (MAdaPT) protocol which formulates the adaptation process as a bi-level optimization framework. To enable scalable meta-training under this framework, we propose a novel heuristic solution, first-order bi-level optimization (FOBLO), avoiding heavy computation costs. Finally, we stabilize meta-training by using a robust initialization through interleaved supervision which alternates self-supervised and supervised objectives. Empirically, SpidR-Adapt achieves rapid gains in phonemic discriminability (ABX) and downstream spoken language modeling scores (sWUGGY, sBLIMP, tSC), surpassing in-domain toplines after training on less than 1h of target-language audio and delivering $100\times$ greater data efficiency than standard multi-task training. These findings highlight a practical, architecture-agnostic path toward biologically inspired, data-efficient representations. We open-source the training code and model checkpoints at https://github.com/facebookresearch/spidr-adapt. --- ### SpidR-Adapt: A Universal Speech Representation Model for Few-Shot Adaptation Authors: Mahi Luthra, Jiayi Shen, Maxime Poli, Angelo Ortiz, Yosuke Higuchi, Youssef Benchekroun, Martin Gleize, Charles-Eric Saint-James, Dongyan Lin, Phillip Rust, Angel Villar, Surya Parimi et al. Published: 2025-12-24 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2512.21204v1 Human infants, with only a few hundred hours of speech exposure, acquire basic units of new languages, highlighting a striking efficiency gap compared to the data-hungry self-supervised speech models. To address this gap, this paper introduces SpidR-Adapt for rapid adaptation to new languages using minimal unlabeled data. We cast such low-resource speech representation learning as a meta-learning problem and construct a multi-task adaptive pre-training (MAdaPT) protocol which formulates the adaptation process as a bi-level optimization framework. To enable scalable meta-training under this framework, we propose a novel heuristic solution, first-order bi-level optimization (FOBLO), avoiding heavy computation costs. Finally, we stabilize meta-training by using a robust initialization through interleaved supervision which alternates self-supervised and supervised objectives. Empirically, SpidR-Adapt achieves rapid gains in phonemic discriminability (ABX) and spoken language modeling (sWUGGY, sBLIMP, tSC), improving over in-domain language models after training on less than 1h of target-language audio, over $100\times$ more data-efficient than standard training. These findings highlight a practical, architecture-agnostic path toward biologically inspired, data-efficient representations. We open-source the training code and model checkpoints at https://github.com/facebookresearch/spidr-adapt. --- ### Schrödinger's Navigator: Imagining an Ensemble of Futures for Zero-Shot Object Navigation Authors: Yu He, Da Huang, Zhenyang Liu, Zixiao Gu, Qiang Sun, Guangnan Ye, Yanwei Fu, Yu-Gang Jiang Published: 2025-12-24 Categories: cs.RO, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2512.21201v4 Zero-shot object navigation (ZSON) requires robots to find target objects in unseen environments without task-specific fine-tuning or pre-built maps, a key capability for general-purpose service robots. Yet methods that perform well in simulation often degrade in cluttered real-world scenes with severe occlusion and latent hazards, where large unseen regions make single-scene inference brittle and unsafe. We propose Schrödinger's Navigator, a belief-aware framework that reasons at inference time over multiple trajectory-conditioned imagined 3D futures. Given candidate paths, a trajectory-conditioned 3D world model predicts hypothetical observations and maintains a superposition of plausible scene realizations rather than committing to one map. An adaptive occluder-aware sampler directs imagination to uncertainty-critical regions, while a Future-Aware Value Map (FAVM) aggregates imagined futures for robust, proactive action selection. Experiments in simulation and on a physical Go2 quadruped show that Schrödinger's Navigator outperforms strong ZSON baselines, improving hidden-target discovery and risk-aware waypoint selection in occlusion-heavy navigation scenarios. These results highlight imagined 3D futures as a scalable and generalizable strategy for zero-shot navigation in uncertain real-world environments. --- ### Schrödinger's Navigator: Imagining an Ensemble of Futures for Zero-Shot Object Navigation Authors: Yu He, Da Huang, Zhenyang Liu, Zixiao Gu, Qiang Sun, Guangnan Ye, Yanwei Fu Published: 2025-12-24 Categories: cs.RO, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2512.21201v1 Zero-shot object navigation (ZSON) requires a robot to locate a target object in a previously unseen environment without relying on pre-built maps or task-specific training. However, existing ZSON methods often struggle in realistic and cluttered environments, particularly when the scene contains heavy occlusions, unknown risks, or dynamically moving target objects. To address these challenges, we propose \textbf{Schrödinger's Navigator}, a navigation framework inspired by Schrödinger's thought experiment on uncertainty. The framework treats unobserved space as a set of plausible future worlds and reasons over them before acting. Conditioned on egocentric visual inputs and three candidate trajectories, a trajectory-conditioned 3D world model imagines future observations along each path. This enables the agent to see beyond occlusions and anticipate risks in unseen regions without requiring extra detours or dense global mapping. The imagined 3D observations are fused into the navigation map and used to update a value map. These updates guide the policy toward trajectories that avoid occlusions, reduce exposure to uncertain space, and better track moving targets. Experiments on a Go2 quadruped robot across three challenging scenarios, including severe static occlusions, unknown risks, and dynamically moving targets, show that Schrödinger's Navigator consistently outperforms strong ZSON baselines in self-localization, object localization, and overall Success Rate in occlusion-heavy environments. These results demonstrate the effectiveness of trajectory-conditioned 3D imagination in enabling robust zero-shot object navigation. --- ### VisRes Bench: On Evaluating the Visual Reasoning Capabilities of VLMs Authors: Brigitta Malagurski Törtei, Yasser Dahou, Ngoc Dung Huynh, Wamiq Reyaz Para, Phúc H. Lê Khac, Ankit Singh, Sofian Chaybouti, Sanath Narayan Published: 2025-12-24 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.21194v1 Vision-Language Models (VLMs) have achieved remarkable progress across tasks such as visual question answering and image captioning. Yet, the extent to which these models perform visual reasoning as opposed to relying on linguistic priors remains unclear. To address this, we introduce VisRes Bench, a benchmark designed to study visual reasoning in naturalistic settings without contextual language supervision. Analyzing model behavior across three levels of complexity, we uncover clear limitations in perceptual and relational visual reasoning capacities. VisRes isolates distinct reasoning abilities across its levels. Level 1 probes perceptual completion and global image matching under perturbations such as blur, texture changes, occlusion, and rotation; Level 2 tests rule-based inference over a single attribute (e.g., color, count, orientation); and Level 3 targets compositional reasoning that requires integrating multiple visual attributes. Across more than 19,000 controlled task images, we find that state-of-the-art VLMs perform near random under subtle perceptual perturbations, revealing limited abstraction beyond pattern recognition. We conclude by discussing how VisRes provides a unified framework for advancing abstract visual reasoning in multimodal research. --- ### UltraShape 1.0: High-Fidelity 3D Shape Generation via Scalable Geometric Refinement Authors: Tanghui Jia, Dongyu Yan, Dehao Hao, Yang Li, Kaiyi Zhang, Xianyi He, Lanjiong Li, Yuhan Wang, Jinnan Chen, Lutao Jiang, Qishen Yin, Long Quan et al. Published: 2025-12-24 Categories: cs.CV, cs.GR arXiv: https://arxiv.org/abs/2512.21185v2 In this report, we introduce UltraShape 1.0, a scalable 3D diffusion framework for high-fidelity 3D geometry generation. The proposed approach adopts a two-stage generation pipeline: a coarse global structure is first synthesized and then refined to produce detailed, high-quality geometry. To support reliable 3D generation, we develop a comprehensive data processing pipeline that includes a novel watertight processing method and high-quality data filtering. This pipeline improves the geometric quality of publicly available 3D datasets by removing low-quality samples, filling holes, and thickening thin structures, while preserving fine-grained geometric details. To enable fine-grained geometry refinement, we decouple spatial localization from geometric detail synthesis in the diffusion process. We achieve this by performing voxel-based refinement at fixed spatial locations, where voxel queries derived from coarse geometry provide explicit positional anchors encoded via RoPE, allowing the diffusion model to focus on synthesizing local geometric details within a reduced, structured solution space. Our model is trained exclusively on publicly available 3D datasets, achieving strong geometric quality despite limited training resources. Extensive evaluations demonstrate that UltraShape 1.0 performs competitively with existing open-source methods in both data processing quality and geometry generation. All code and trained models will be released to support future research. --- ### UltraShape 1.0: High-Fidelity 3D Shape Generation via Scalable Geometric Refinement Authors: Tanghui Jia, Dongyu Yan, Dehao Hao, Yang Li, Kaiyi Zhang, Xianyi He, Lanjiong Li, Jinnan Chen, Lutao Jiang, Qishen Yin, Long Quan, Ying-Cong Chen et al. Published: 2025-12-24 Categories: cs.CV, cs.GR arXiv: https://arxiv.org/abs/2512.21185v1 In this report, we introduce UltraShape 1.0, a scalable 3D diffusion framework for high-fidelity 3D geometry generation. The proposed approach adopts a two-stage generation pipeline: a coarse global structure is first synthesized and then refined to produce detailed, high-quality geometry. To support reliable 3D generation, we develop a comprehensive data processing pipeline that includes a novel watertight processing method and high-quality data filtering. This pipeline improves the geometric quality of publicly available 3D datasets by removing low-quality samples, filling holes, and thickening thin structures, while preserving fine-grained geometric details. To enable fine-grained geometry refinement, we decouple spatial localization from geometric detail synthesis in the diffusion process. We achieve this by performing voxel-based refinement at fixed spatial locations, where voxel queries derived from coarse geometry provide explicit positional anchors encoded via RoPE, allowing the diffusion model to focus on synthesizing local geometric details within a reduced, structured solution space. Our model is trained exclusively on publicly available 3D datasets, achieving strong geometric quality despite limited training resources. Extensive evaluations demonstrate that UltraShape 1.0 performs competitively with existing open-source methods in both data processing quality and geometry generation. All code and trained models will be released to support future research. --- ### Towards Arbitrary Motion Completing via Hierarchical Continuous Representation Authors: Chenghao Xu, Guangtao Lyu, Qi Liu, Jiexi Yan, Muli Yang, Cheng Deng Published: 2025-12-24 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.21183v1 Physical motions are inherently continuous, and higher camera frame rates typically contribute to improved smoothness and temporal coherence. For the first time, we explore continuous representations of human motion sequences, featuring the ability to interpolate, inbetween, and even extrapolate any input motion sequences at arbitrary frame rates. To achieve this, we propose a novel parametric activation-induced hierarchical implicit representation framework, referred to as NAME, based on Implicit Neural Representations (INRs). Our method introduces a hierarchical temporal encoding mechanism that extracts features from motion sequences at multiple temporal scales, enabling effective capture of intricate temporal patterns. Additionally, we integrate a custom parametric activation function, powered by Fourier transformations, into the MLP-based decoder to enhance the expressiveness of the continuous representation. This parametric formulation significantly augments the model's ability to represent complex motion behaviors with high accuracy. Extensive evaluations across several benchmark datasets demonstrate the effectiveness and robustness of our proposed approach. --- ### Equivariant Multiscale Learned Invertible Reconstruction for Cone Beam CT: From Simulated to Real Data Authors: Nikita Moriakov, Efstratios Gavves, Jonathan H. Mason, Carmen Seller-Oria, Jonas Teuwen, Jan-Jakob Sonke Published: 2025-12-24 Categories: physics.med-ph, cs.CV arXiv: https://arxiv.org/abs/2512.21180v1 Cone Beam CT (CBCT) is an important imaging modality nowadays, however lower image quality of CBCT compared to more conventional Computed Tomography (CT) remains a limiting factor in CBCT applications. Deep learning reconstruction methods are a promising alternative to classical analytical and iterative reconstruction methods, but applying such methods to CBCT is often difficult due to the lack of ground truth data, memory limitations and the need for fast inference at clinically-relevant resolutions. In this work we propose LIRE++, an end-to-end rotationally-equivariant multiscale learned invertible primal-dual scheme for fast and memory-efficient CBCT reconstruction. Memory optimizations and multiscale reconstruction allow for fast training and inference, while rotational equivariance improves parameter efficiency. LIRE++ was trained on simulated projection data from a fast quasi-Monte Carlo CBCT projection simulator that we developed as well. Evaluated on synthetic data, LIRE++ gave an average improvement of 1 dB in Peak Signal-to-Noise Ratio over alternative deep learning baselines. On real clinical data, LIRE++ improved the average Mean Absolute Error between the reconstruction and the corresponding planning CT by 10 Hounsfield Units with respect to current proprietary state-of-the-art hybrid deep-learning/iterative method. --- ### A Turn Toward Better Alignment: Few-Shot Generative Adaptation with Equivariant Feature Rotation Authors: Chenghao Xu, Qi Liu, Jiexi Yan, Muli Yang, Cheng Deng Published: 2025-12-24 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.21174v1 Few-shot image generation aims to effectively adapt a source generative model to a target domain using very few training images. Most existing approaches introduce consistency constraints-typically through instance-level or distribution-level loss functions-to directly align the distribution patterns of source and target domains within their respective latent spaces. However, these strategies often fall short: overly strict constraints can amplify the negative effects of the domain gap, leading to distorted or uninformative content, while overly relaxed constraints may fail to leverage the source domain effectively. This limitation primarily stems from the inherent discrepancy in the underlying distribution structures of the source and target domains. The scarcity of target samples further compounds this issue by hindering accurate estimation of the target domain's distribution. To overcome these limitations, we propose Equivariant Feature Rotation (EFR), a novel adaptation strategy that aligns source and target domains at two complementary levels within a self-rotated proxy feature space. Specifically, we perform adaptive rotations within a parameterized Lie Group to transform both source and target features into an equivariant proxy space, where alignment is conducted. These learnable rotation matrices serve to bridge the domain gap by preserving intra-domain structural information without distortion, while the alignment optimization facilitates effective knowledge transfer from the source to the target domain. Comprehensive experiments on a variety of commonly used datasets demonstrate that our method significantly enhances the generative performance within the targeted domain. --- ### A Unified Framework for EEG Seizure Detection Using Universum-Integrated Generalized Eigenvalues Proximal Support Vector Machine Authors: Yogesh Kumar, Vrushank Ahire, M. A. Ganaie Published: 2025-12-24 Categories: cs.LG arXiv: https://arxiv.org/abs/2512.21170v1 The paper presents novel Universum-enhanced classifiers: the Universum Generalized Eigenvalue Proximal Support Vector Machine (U-GEPSVM) and the Improved U-GEPSVM (IU-GEPSVM) for EEG signal classification. Using the computational efficiency of generalized eigenvalue decomposition and the generalization benefits of Universum learning, the proposed models address critical challenges in EEG analysis: non-stationarity, low signal-to-noise ratio, and limited labeled data. U-GEPSVM extends the GEPSVM framework by incorporating Universum constraints through a ratio-based objective function, while IU-GEPSVM enhances stability through a weighted difference-based formulation that provides independent control over class separation and Universum alignment. The models are evaluated on the Bonn University EEG dataset across two binary classification tasks: (O vs S)-healthy (eyes closed) vs seizure, and (Z vs S)-healthy (eyes open) vs seizure. IU-GEPSVM achieves peak accuracies of 85% (O vs S) and 80% (Z vs S), with mean accuracies of 81.29% and 77.57% respectively, outperforming baseline methods. --- ### A Community-Enhanced Graph Representation Model for Link Prediction Authors: Lei Wang, Darong Lai Published: 2025-12-24 Categories: cs.SI, cs.LG arXiv: https://arxiv.org/abs/2512.21166v1 Although Graph Neural Networks (GNNs) have become the dominant approach for graph representation learning, their performance on link prediction tasks does not always surpass that of traditional heuristic methods such as Common Neighbors and Jaccard Coefficient. This is mainly because existing GNNs tend to focus on learning local node representations, making it difficult to effectively capture structural relationships between node pairs. Furthermore, excessive reliance on local neighborhood information can lead to over-smoothing. Prior studies have shown that introducing global structural encoding can partially alleviate this issue. To address these limitations, we propose a Community-Enhanced Link Prediction (CELP) framework that incorporates community structure to jointly model local and global graph topology. Specifically, CELP enhances the graph via community-aware, confidence-guided edge completion and pruning, while integrating multi-scale structural features to achieve more accurate link prediction. Experimental results across multiple benchmark datasets demonstrate that CELP achieves superior performance, validating the crucial role of community structure in improving link prediction accuracy. --- ### BALLAST: Bandit-Assisted Learning for Latency-Aware Stable Timeouts in Raft Authors: Qizhi Wang Published: 2025-12-24 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2512.21165v1 Randomized election timeouts are a simple and effective liveness heuristic for Raft, but they become brittle under long-tail latency, jitter, and partition recovery, where repeated split votes can inflate unavailability. This paper presents BALLAST, a lightweight online adaptation mechanism that replaces static timeout heuristics with contextual bandits. BALLAST selects from a discrete set of timeout "arms" using efficient linear contextual bandits (LinUCB variants), and augments learning with safe exploration to cap risk during unstable periods. We evaluate BALLAST on a reproducible discrete-event simulation with long-tail delay, loss, correlated bursts, node heterogeneity, and partition/recovery turbulence. Across challenging WAN regimes, BALLAST substantially reduces recovery time and unwritable time compared to standard randomized timeouts and common heuristics, while remaining competitive on stable LAN/WAN settings. --- ### ElfCore: A 28nm Neural Processor Enabling Dynamic Structured Sparse Training and Online Self-Supervised Learning with Activity-Dependent Weight Update Authors: Zhe Su, Giacomo Indiveri Published: 2025-12-24 Categories: cs.AR, cs.LG arXiv: https://arxiv.org/abs/2512.21153v1 In this paper, we present ElfCore, a 28nm digital spiking neural network processor tailored for event-driven sensory signal processing. ElfCore is the first to efficiently integrate: (1) a local online self-supervised learning engine that enables multi-layer temporal learning without labeled inputs; (2) a dynamic structured sparse training engine that supports high-accuracy sparse-to-sparse learning; and (3) an activity-dependent sparse weight update mechanism that selectively updates weights based solely on input activity and network dynamics. Demonstrated on tasks including gesture recognition, speech, and biomedical signal processing, ElfCore outperforms state-of-the-art solutions with up to 16X lower power consumption, 3.8X reduced on-chip memory requirements, and 5.9X greater network capacity efficiency. --- ### MODE: Multi-Objective Adaptive Coreset Selection Authors: Tanmoy Mukherjee, Pierre Marquis, Zied Bouraoui Published: 2025-12-24 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2512.21152v1 We present Mode(Multi-Objective adaptive Data Efficiency), a framework that dynamically combines coreset selection strategies based on their evolving contribution to model performance. Unlike static methods, \mode adapts selection criteria to training phases: emphasizing class balance early, diversity during representation learning, and uncertainty at convergence. We show that MODE achieves (1-1/e)-approximation with O(n \log n) complexity and demonstrates competitive accuracy while providing interpretable insights into data utility evolution. Experiments show \mode reduces memory requirements --- ### ORCA: Object Recognition and Comprehension for Archiving Marine Species Authors: Yuk-Kwan Wong, Haixin Liang, Zeyu Ma, Yiwei Chen, Ziqiang Zheng, Rinaldi Gotama, Pascal Sebastian, Lauren D. Sparks, Sai-Kit Yeung Published: 2025-12-24 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.21150v1 Marine visual understanding is essential for monitoring and protecting marine ecosystems, enabling automatic and scalable biological surveys. However, progress is hindered by limited training data and the lack of a systematic task formulation that aligns domain-specific marine challenges with well-defined computer vision tasks, thereby limiting effective model application. To address this gap, we present ORCA, a multi-modal benchmark for marine research comprising 14,647 images from 478 species, with 42,217 bounding box annotations and 22,321 expert-verified instance captions. The dataset provides fine-grained visual and textual annotations that capture morphology-oriented attributes across diverse marine species. To catalyze methodological advances, we evaluate 18 state-of-the-art models on three tasks: object detection (closed-set and open-vocabulary), instance captioning, and visual grounding. Results highlight key challenges, including species diversity, morphological overlap, and specialized domain demands, underscoring the difficulty of marine understanding. ORCA thus establishes a comprehensive benchmark to advance research in marine domain. Project Page: http://orca.hkustvgd.com/. --- ### TGC-Net: A Structure-Aware and Semantically-Aligned Framework for Text-Guided Medical Image Segmentation Authors: Gaoren Lin, Huangxuan Zhao, Yuan Xiong, Lefei Zhang, Bo Du, Wentao Zhu Published: 2025-12-24 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2512.21135v1 Text-guided medical segmentation enhances segmentation accuracy by utilizing clinical reports as auxiliary information. However, existing methods typically rely on unaligned image and text encoders, which necessitate complex interaction modules for multimodal fusion. While CLIP provides a pre-aligned multimodal feature space, its direct application to medical imaging is limited by three main issues: insufficient preservation of fine-grained anatomical structures, inadequate modeling of complex clinical descriptions, and domain-specific semantic misalignment. To tackle these challenges, we propose TGC-Net, a CLIP-based framework focusing on parameter-efficient, task-specific adaptations. Specifically, it incorporates a Semantic-Structural Synergy Encoder (SSE) that augments CLIP's ViT with a CNN branch for multi-scale structural refinement, a Domain-Augmented Text Encoder (DATE) that injects large-language-model-derived medical knowledge, and a Vision-Language Calibration Module (VLCM) that refines cross-modal correspondence in a unified feature space. Experiments on five datasets across chest X-ray and thoracic CT modalities demonstrate that TGC-Net achieves state-of-the-art performance with substantially fewer trainable parameters, including notable Dice gains on challenging benchmarks. --- ### AutoBaxBuilder: Bootstrapping Code Security Benchmarking Authors: Tobias von Arx, Niels Mündler, Mark Vero, Maximilian Baader, Martin Vechev Published: 2025-12-24 Categories: cs.CR, cs.AI, cs.LG, cs.PL arXiv: https://arxiv.org/abs/2512.21132v2 As large language models (LLMs) see wide adoption in software engineering, the reliable assessment of the correctness and security of LLM-generated code is crucial. Notably, prior work showed that LLMs are prone to generating code with security vulnerabilities, highlighting that security is often overlooked. These insights were enabled by specialized benchmarks crafted by security experts through significant manual effort. However, benchmarks (i) inevitably end up contaminating training data, (ii) must extend to new tasks to provide a more complete picture, and (iii) must increase in difficulty to challenge more capable LLMs. In this work, we address these challenges and present AutoBaxBuilder, an automated pipeline that generates code security benchmarking tasks from scratch. It leverages the code-understanding capabilities of LLMs combined with robust reliability checks to construct functional tests and end-to-end security-probing exploits. The quality of the pipeline is quantitatively confirmed by aligning its predictions with an expert-written baseline and qualitatively validated through manual soundness verification. We use AutoBaxBuilder to construct a new benchmark and release it to the public as AutoBaxBench, together with a thorough evaluation on contemporary LLMs. AutoBaxBuilder generates new tasks in under 2 hours, for less than USD 4. Including a manual verification, this reduces the required human effort for benchmark construction by a factor of 12. --- ### AutoBaxBuilder: Bootstrapping Code Security Benchmarking Authors: Tobias von Arx, Niels Mündler, Mark Vero, Maximilian Baader, Martin Vechev Published: 2025-12-24 Categories: cs.CR, cs.AI, cs.LG, cs.PL arXiv: https://arxiv.org/abs/2512.21132v1 As LLMs see wide adoption in software engineering, the reliable assessment of the correctness and security of LLM-generated code is crucial. Notably, prior work has demonstrated that security is often overlooked, exposing that LLMs are prone to generating code with security vulnerabilities. These insights were enabled by specialized benchmarks, crafted through significant manual effort by security experts. However, relying on manually-crafted benchmarks is insufficient in the long term, because benchmarks (i) naturally end up contaminating training data, (ii) must extend to new tasks to provide a more complete picture, and (iii) must increase in difficulty to challenge more capable LLMs. In this work, we address these challenges and present AutoBaxBuilder, a framework that generates tasks and tests for code security benchmarking from scratch. We introduce a robust pipeline with fine-grained plausibility checks, leveraging the code understanding capabilities of LLMs to construct functionality tests and end-to-end security-probing exploits. To confirm the quality of the generated benchmark, we conduct both a qualitative analysis and perform quantitative experiments, comparing it against tasks constructed by human experts. We use AutoBaxBuilder to construct entirely new tasks and release them to the public as AutoBaxBench, together with a thorough evaluation of the security capabilities of LLMs on these tasks. We find that a new task can be generated in under 2 hours, costing less than USD 10. --- ### Active inference and artificial reasoning Authors: Karl Friston, Lancelot Da Costa, Alexander Tschantz, Conor Heins, Christopher Buckley, Tim Verbelen, Thomas Parr Published: 2025-12-24 Categories: q-bio.NC, physics.data-an, stat.ML arXiv: https://arxiv.org/abs/2512.21129v1 This technical note considers the sampling of outcomes that provide the greatest amount of information about the structure of underlying world models. This generalisation furnishes a principled approach to structure learning under a plausible set of generative models or hypotheses. In active inference, policies - i.e., combinations of actions - are selected based on their expected free energy, which comprises expected information gain and value. Information gain corresponds to the KL divergence between predictive posteriors with, and without, the consequences of action. Posteriors over models can be evaluated quickly and efficiently using Bayesian Model Reduction, based upon accumulated posterior beliefs about model parameters. The ensuing information gain can then be used to select actions that disambiguate among alternative models, in the spirit of optimal experimental design. We illustrate this kind of active selection or reasoning using partially observed discrete models; namely, a 'three-ball' paradigm used previously to describe artificial insight and 'aha moments' via (synthetic) introspection or sleep. We focus on the sample efficiency afforded by seeking outcomes that resolve the greatest uncertainty about the world model, under which outcomes are generated. --- ### A Real-World Evaluation of LLM Medication Safety Reviews in NHS Primary Care Authors: Oliver Normand, Esther Borsi, Mitch Fruin, Lauren E Walker, Jamie Heagerty, Chris C. Holmes, Anthony J Avery, Iain E Buchan, Harry Coppock Published: 2025-12-24 Categories: cs.AI arXiv: https://arxiv.org/abs/2512.21127v1 Large language models (LLMs) often match or exceed clinician-level performance on medical benchmarks, yet very few are evaluated on real clinical data or examined beyond headline metrics. We present, to our knowledge, the first evaluation of an LLM-based medication safety review system on real NHS primary care data, with detailed characterisation of key failure behaviours across varying levels of clinical complexity. In a retrospective study using a population-scale EHR spanning 2,125,549 adults in NHS Cheshire and Merseyside, we strategically sampled patients to capture a broad range of clinical complexity and medication safety risk, yielding 277 patients after data-quality exclusions. An expert clinician reviewed these patients and graded system-identified issues and proposed interventions. Our primary LLM system showed strong performance in recognising when a clinical issue is present (sensitivity 100\% (95\% CI 98.2--100), specificity 83.1\% (95\% CI 72.7--90.1)), yet correctly identified all issues and interventions in only 46.9\% (95\% CI 41.1--52.8) of patients. Failure analysis reveals that, in this setting, the dominant failure mechanism is contextual reasoning rather than missing medication knowledge, with five primary patterns: overconfidence in uncertainty, applying standard guidelines without adjusting for patient context, misunderstanding how healthcare is delivered in practice, factual errors, and process blindness. These patterns persisted across patient complexity and demographic strata, and across a range of state-of-the-art models and configurations. We provide 45 detailed vignettes that comprehensively cover all identified failure cases. This work highlights shortcomings that must be addressed before LLM-based clinical AI can be safely deployed. It also begs larger-scale, prospective evaluations and deeper study of LLM behaviours in clinical contexts. --- ### MarineEval: Assessing the Marine Intelligence of Vision-Language Models Authors: YuK-Kwan Wong, Tuan-An To, Jipeng Zhang, Ziqiang Zheng, Sai-Kit Yeung Published: 2025-12-24 Categories: cs.CV, cs.DB arXiv: https://arxiv.org/abs/2512.21126v1 We have witnessed promising progress led by large language models (LLMs) and further vision language models (VLMs) in handling various queries as a general-purpose assistant. VLMs, as a bridge to connect the visual world and language corpus, receive both visual content and various text-only user instructions to generate corresponding responses. Though great success has been achieved by VLMs in various fields, in this work, we ask whether the existing VLMs can act as domain experts, accurately answering marine questions, which require significant domain expertise and address special domain challenges/requirements. To comprehensively evaluate the effectiveness and explore the boundary of existing VLMs, we construct the first large-scale marine VLM dataset and benchmark called MarineEval, with 2,000 image-based question-answering pairs. During our dataset construction, we ensure the diversity and coverage of the constructed data: 7 task dimensions and 20 capacity dimensions. The domain requirements are specially integrated into the data construction and further verified by the corresponding marine domain experts. We comprehensively benchmark 17 existing VLMs on our MarineEval and also investigate the limitations of existing models in answering marine research questions. The experimental results reveal that existing VLMs cannot effectively answer the domain-specific questions, and there is still a large room for further performance improvements. We hope our new benchmark and observations will facilitate future research. Project Page: http://marineeval.hkustvgd.com/ --- ### ClarifyMT-Bench: Benchmarking and Improving Multi-Turn Clarification for Conversational Large Language Models Authors: Sichun Luo, Yi Huang, Mukai Li, Shichang Meng, Fengyuan Liu, Zefa Hu, Junlan Feng, Qi Liu Published: 2025-12-24 Categories: cs.CL, cs.IR arXiv: https://arxiv.org/abs/2512.21120v1 Large language models (LLMs) are increasingly deployed as conversational assistants in open-domain, multi-turn settings, where users often provide incomplete or ambiguous information. However, existing LLM-focused clarification benchmarks primarily assume single-turn interactions or cooperative users, limiting their ability to evaluate clarification behavior in realistic settings. We introduce \textbf{ClarifyMT-Bench}, a benchmark for multi-turn clarification grounded in a five-dimensional ambiguity taxonomy and a set of six behaviorally diverse simulated user personas. Through a hybrid LLM-human pipeline, we construct 6,120 multi-turn dialogues capturing diverse ambiguity sources and interaction patterns. Evaluating ten representative LLMs uncovers a consistent under-clarification bias: LLMs tend to answer prematurely, and performance degrades as dialogue depth increases. To mitigate this, we propose \textbf{ClarifyAgent}, an agentic approach that decomposes clarification into perception, forecasting, tracking, and planning, substantially improving robustness across ambiguity conditions. ClarifyMT-Bench establishes a reproducible foundation for studying when LLMs should ask, when they should answer, and how to navigate ambiguity in real-world human-LLM interactions. --- ### STLDM: Spatio-Temporal Latent Diffusion Model for Precipitation Nowcasting Authors: Shi Quan Foo, Chi-Ho Wong, Zhihan Gao, Dit-Yan Yeung, Ka-Hing Wong, Wai-Kin Wong Published: 2025-12-24 Categories: cs.LG, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2512.21118v1 Precipitation nowcasting is a critical spatio-temporal prediction task for society to prevent severe damage owing to extreme weather events. Despite the advances in this field, the complex and stochastic nature of this task still poses challenges to existing approaches. Specifically, deterministic models tend to produce blurry predictions while generative models often struggle with poor accuracy. In this paper, we present a simple yet effective model architecture termed STLDM, a diffusion-based model that learns the latent representation from end to end alongside both the Variational Autoencoder and the conditioning network. STLDM decomposes this task into two stages: a deterministic forecasting stage handled by the conditioning network, and an enhancement stage performed by the latent diffusion model. Experimental results on multiple radar datasets demonstrate that STLDM achieves superior performance compared to the state of the art, while also improving inference efficiency. The code is available in https://github.com/sqfoo/stldm_official. --- ### A Mechanistic Analysis of Transformers for Dynamical Systems Authors: Gregory Duthé, Nikolaos Evangelou, Wei Liu, Ioannis G. Kevrekidis, Eleni Chatzi Published: 2025-12-24 Categories: cs.LG, cs.CE arXiv: https://arxiv.org/abs/2512.21113v2 Transformers are increasingly adopted for modeling and forecasting time-series, yet their internal mechanisms remain poorly understood from a dynamical systems perspective. In contrast to classical autoregressive and state-space models, which benefit from well-established theoretical foundations, Transformer architectures are typically treated as black boxes. This gap becomes particularly relevant as attention-based models are considered for general-purpose or zero-shot forecasting across diverse dynamical regimes. In this work, we do not propose a new forecasting model, but instead investigate the representational capabilities and limitations of single-layer Transformers when applied to dynamical data. Building on a dynamical systems perspective, we interpret causal self-attention as a linear, history-dependent recurrence and analyze how it processes temporal information. Through a series of linear and nonlinear case studies, we identify distinct operational regimes. For linear systems, we show that in the single-head attention-only setting, the convexity constraint imposed by softmax attention restricts the class of autoregressive operators that can be represented, leading to oversmoothing when the target dynamics require mixed-sign lag coefficients. For nonlinear systems under partial observability, attention instead acts as an adaptive delay-embedding mechanism, enabling effective state reconstruction when sufficient temporal context and latent dimensionality are available. These results help bridge empirical observations with classical dynamical systems theory, providing insight into when and why Transformers succeed or fail as models of dynamical systems. --- ### A Mechanistic Analysis of Transformers for Dynamical Systems Authors: Gregory Duthé, Nikolaos Evangelou, Wei Liu, Ioannis G. Kevrekidis, Eleni Chatzi Published: 2025-12-24 Categories: cs.LG, cs.CE arXiv: https://arxiv.org/abs/2512.21113v1 Transformers are increasingly adopted for modeling and forecasting time-series, yet their internal mechanisms remain poorly understood from a dynamical systems perspective. In contrast to classical autoregressive and state-space models, which benefit from well-established theoretical foundations, Transformer architectures are typically treated as black boxes. This gap becomes particularly relevant as attention-based models are considered for general-purpose or zero-shot forecasting across diverse dynamical regimes. In this work, we do not propose a new forecasting model, but instead investigate the representational capabilities and limitations of single-layer Transformers when applied to dynamical data. Building on a dynamical systems perspective we interpret causal self-attention as a linear, history-dependent recurrence and analyze how it processes temporal information. Through a series of linear and nonlinear case studies, we identify distinct operational regimes. For linear systems, we show that the convexity constraint imposed by softmax attention fundamentally restricts the class of dynamics that can be represented, leading to oversmoothing in oscillatory settings. For nonlinear systems under partial observability, attention instead acts as an adaptive delay-embedding mechanism, enabling effective state reconstruction when sufficient temporal context and latent dimensionality are available. These results help bridge empirical observations with classical dynamical systems theory, providing insight into when and why Transformers succeed or fail as models of dynamical systems. --- ### Statistical and computational challenges in ranking Authors: Alexandra Carpentier, Nicolas Verzelen Published: 2025-12-24 Categories: math.ST, stat.ML arXiv: https://arxiv.org/abs/2512.21111v1 We consider the problem of ranking $n$ experts according to their abilities, based on the correctness of their answers to $d$ questions. This is modeled by the so-called crowd-sourcing model, where the answer of expert $i$ on question $k$ is modeled by a random entry, parametrized by $M_{i,k}$ which is increasing linearly with the expected quality of the answer. To enable the unambiguous ranking of the experts by ability, several assumptions on $M$ are available in the literature. We consider here the general isotonic crowd-sourcing model, where $M$ is assumed to be isotonic up to an unknown permutation $π^*$ of the experts - namely, $M_{π^{*-1}(i),k} \geq M_{π^{*-1}(i+1),k}$ for any $i\in (n-1), k \in (d)$. Then, ranking experts amounts to constructing an estimator of $π^*$. In particular, we investigate here the existence of statistically optimal and computationally efficient procedures and we describe recent results that disprove the existence of computational-statistical gaps for this problem. To provide insights on the key ideas, we start by discussing simpler and yet related sub-problems, namely sub-matrix detection and estimation. This corresponds to specific instances of the ranking problem where the matrix $M$ is constrained to be of the form $λ\mathbf 1\{S\times T\}$ where $S\subset (n), T\subset (d)$. This model has been extensively studied. We provide an overview of the results and proof techniques for this problem with a particular emphasis on the computational lower bounds based on low-degree polynomial methods. Then, we build upon this instrumental sub-problem to discuss existing results and algorithmic ideas for the general ranking problem. --- ### Beyond Context: Large Language Models' Failure to Grasp Users' Intent Authors: Ahmed M. Hussain, Salahuddin Salahuddin Published: 2025-12-24 Categories: cs.AI, cs.CL, cs.CR, cs.CY arXiv: https://arxiv.org/abs/2512.21110v3 Current Large Language Models (LLMs) safety approaches focus on explicitly harmful content while overlooking a critical vulnerability: the inability to understand context and recognize user intent. This creates exploitable vulnerabilities that malicious users can systematically leverage to circumvent safety mechanisms. We empirically evaluate multiple state-of-the-art LLMs, including ChatGPT, Claude, Gemini, and DeepSeek. Our analysis demonstrates the circumvention of reliable safety mechanisms through emotional framing, progressive revelation, and academic justification techniques. Notably, reasoning-enabled configurations amplified rather than mitigated the effectiveness of exploitation, increasing factual precision while failing to interrogate the underlying intent. The exception was Claude Opus 4.1, which prioritized intent detection over information provision in some use cases. This pattern reveals that current architectural designs create systematic vulnerabilities. These limitations require paradigmatic shifts toward contextual understanding and intent recognition as core safety capabilities rather than post-hoc protective mechanisms. --- ### Beyond Context: Large Language Models Failure to Grasp Users Intent Authors: Ahmed M. Hussain, Salahuddin Salahuddin, Panos Papadimitratos Published: 2025-12-24 Categories: cs.AI, cs.CL, cs.CR, cs.CY arXiv: https://arxiv.org/abs/2512.21110v1 Current Large Language Models (LLMs) safety approaches focus on explicitly harmful content while overlooking a critical vulnerability: the inability to understand context and recognize user intent. This creates exploitable vulnerabilities that malicious users can systematically leverage to circumvent safety mechanisms. We empirically evaluate multiple state-of-the-art LLMs, including ChatGPT, Claude, Gemini, and DeepSeek. Our analysis demonstrates the circumvention of reliable safety mechanisms through emotional framing, progressive revelation, and academic justification techniques. Notably, reasoning-enabled configurations amplified rather than mitigated the effectiveness of exploitation, increasing factual precision while failing to interrogate the underlying intent. The exception was Claude Opus 4.1, which prioritized intent detection over information provision in some use cases. This pattern reveals that current architectural designs create systematic vulnerabilities. These limitations require paradigmatic shifts toward contextual understanding and intent recognition as core safety capabilities rather than post-hoc protective mechanisms. --- ### Semi-Supervised Learning for Large Language Models Safety and Content Moderation Authors: Eduard Stefan Dinuta, Iustin Sirbu, Traian Rebedea Published: 2025-12-24 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2512.21107v1 Safety for Large Language Models (LLMs) has been an ongoing research focus since their emergence and is even more relevant nowadays with the increasing capacity of those models. Currently, there are several guardrails in place for all public LLMs and multiple proposed datasets for training safety classifiers. However, training these safety classifiers relies on large quantities of labeled data, which can be problematic to acquire, prone to labeling errors, or often include synthetic data. To address these issues, we suggest a different approach: utilizing semi-supervised learning techniques, which leverage both labeled and unlabeled data, to improve the performance on the safety task. We analyze the improvements that these techniques can offer for both prompts given to Large Language Models and the responses to those requests. Moreover, since augmentation is the central part of semi-supervised algorithms, we demonstrate the importance of using task-specific augmentations, which significantly increase the performance when compared to general-purpose augmentation techniques. --- ### Semantic Refinement with LLMs for Graph Representations Authors: Safal Thapaliya, Zehong Wang, Jiazheng Li, Ziming Li, Yanfang Ye, Chuxu Zhang Published: 2025-12-24 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2512.21106v2 Graph-structured data exhibit substantial heterogeneity in where their predictive signals originate: in some domains, node-level semantics dominate, while in others, structural patterns play a central role. This structure-semantics heterogeneity implies that no graph learning model with a fixed inductive bias can generalize optimally across diverse graph domains. However, most existing methods address this challenge from the model side by incrementally injecting new inductive biases, which remains fundamentally limited given the open-ended diversity of real-world graphs. In this work, we take a data-centric perspective and treat node semantics as a task-adaptive variable. We propose a Graph-Exemplar-guided Semantic Refinement (GES) framework for graph representation learning which -- unlike existing LLM-enhanced methods that generate node descriptions without graph context -- leverages structurally and semantically similar nodes from the graph itself to guide semantic refinement. Specifically, a GNN is first trained to produce predictive states, which along with structural and semantic similarity are used to retrieve in-graph exemplars that inform an LLM in refining node descriptions. We evaluate our approach on both text-rich and text-free graphs. Results show consistent improvements on semantics-rich and structure-dominated graphs, demonstrating the effectiveness of data-centric semantic refinement under structure-semantics heterogeneity. --- ### Semantic Refinement with LLMs for Graph Representations Authors: Safal Thapaliya, Zehong Wang, Jiazheng Li, Ziming Li, Yanfang Ye, Chuxu Zhang Published: 2025-12-24 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2512.21106v1 Graph-structured data exhibit substantial heterogeneity in where their predictive signals originate: in some domains, node-level semantics dominate, while in others, structural patterns play a central role. This structure-semantics heterogeneity implies that no graph learning model with a fixed inductive bias can generalize optimally across diverse graph domains. However, most existing methods address this challenge from the model side by incrementally injecting new inductive biases, which remains fundamentally limited given the open-ended diversity of real-world graphs. In this work, we take a data-centric perspective and treat node semantics as a task-adaptive variable. We propose a Data-Adaptive Semantic Refinement framework DAS for graph representation learning, which couples a fixed graph neural network (GNN) and a large language model (LLM) in a closed feedback loop. The GNN provides implicit supervisory signals to guide the semantic refinement of LLM, and the refined semantics are fed back to update the same graph learner. We evaluate our approach on both text-rich and text-free graphs. Results show consistent improvements on structure-dominated graphs while remaining competitive on semantics-rich graphs, demonstrating the effectiveness of data-centric semantic adaptation under structure-semantics heterogeneity. --- ### FreeInpaint: Tuning-free Prompt Alignment and Visual Rationality Enhancement in Image Inpainting Authors: Chao Gong, Dong Li, Yingwei Pan, Jingjing Chen, Ting Yao, Tao Mei Published: 2025-12-24 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.21104v1 Text-guided image inpainting endeavors to generate new content within specified regions of images using textual prompts from users. The primary challenge is to accurately align the inpainted areas with the user-provided prompts while maintaining a high degree of visual fidelity. While existing inpainting methods have produced visually convincing results by leveraging the pre-trained text-to-image diffusion models, they still struggle to uphold both prompt alignment and visual rationality simultaneously. In this work, we introduce FreeInpaint, a plug-and-play tuning-free approach that directly optimizes the diffusion latents on the fly during inference to improve the faithfulness of the generated images. Technically, we introduce a prior-guided noise optimization method that steers model attention towards valid inpainting regions by optimizing the initial noise. Furthermore, we meticulously design a composite guidance objective tailored specifically for the inpainting task. This objective efficiently directs the denoising process, enhancing prompt alignment and visual rationality by optimizing intermediate latents at each step. Through extensive experiments involving various inpainting diffusion models and evaluation metrics, we demonstrate the effectiveness and robustness of our proposed FreeInpaint. --- ### Shared Representation Learning for High-Dimensional Multi-Task Forecasting under Resource Contention in Cloud-Native Backends Authors: Zixiao Huang, Jixiao Yang, Sijia Li, Chi Zhang, Jinyu Chen, Chengda Xu Published: 2025-12-24 Categories: cs.LG arXiv: https://arxiv.org/abs/2512.21102v1 This study proposes a unified forecasting framework for high-dimensional multi-task time series to meet the prediction demands of cloud native backend systems operating under highly dynamic loads, coupled metrics, and parallel tasks. The method builds a shared encoding structure to represent diverse monitoring indicators in a unified manner and employs a state fusion mechanism to capture trend changes and local disturbances across different time scales. A cross-task structural propagation module is introduced to model potential dependencies among nodes, enabling the model to understand complex structural patterns formed by resource contention, link interactions, and changes in service topology. To enhance adaptability to non-stationary behaviors, the framework incorporates a dynamic adjustment mechanism that automatically regulates internal feature flows according to system state changes, ensuring stable predictions in the presence of sudden load shifts, topology drift, and resource jitter. The experimental evaluation compares multiple models across various metrics and verifies the effectiveness of the framework through analyses of hyperparameter sensitivity, environmental sensitivity, and data sensitivity. The results show that the proposed method achieves superior performance on several error metrics and provides more accurate representations of future states under different operating conditions. Overall, the unified forecasting framework offers reliable predictive capability for high-dimensional, multi-task, and strongly dynamic environments in cloud native systems and provides essential technical support for intelligent backend management. --- ### TexAvatars : Hybrid Texel-3D Representations for Stable Rigging of Photorealistic Gaussian Head Avatars Authors: Jaeseong Lee, Junyeong Ahn, Taewoong Kang, Jaegul Choo Published: 2025-12-24 Categories: cs.GR, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2512.21099v1 Constructing drivable and photorealistic 3D head avatars has become a central task in AR/XR, enabling immersive and expressive user experiences. With the emergence of high-fidelity and efficient representations such as 3D Gaussians, recent works have pushed toward ultra-detailed head avatars. Existing approaches typically fall into two categories: rule-based analytic rigging or neural network-based deformation fields. While effective in constrained settings, both approaches often fail to generalize to unseen expressions and poses, particularly in extreme reenactment scenarios. Other methods constrain Gaussians to the global texel space of 3DMMs to reduce rendering complexity. However, these texel-based avatars tend to underutilize the underlying mesh structure. They apply minimal analytic deformation and rely heavily on neural regressors and heuristic regularization in UV space, which weakens geometric consistency and limits extrapolation to complex, out-of-distribution deformations. To address these limitations, we introduce TexAvatars, a hybrid avatar representation that combines the explicit geometric grounding of analytic rigging with the spatial continuity of texel space. Our approach predicts local geometric attributes in UV space via CNNs, but drives 3D deformation through mesh-aware Jacobians, enabling smooth and semantically meaningful transitions across triangle boundaries. This hybrid design separates semantic modeling from geometric control, resulting in improved generalization, interpretability, and stability. Furthermore, TexAvatars captures fine-grained expression effects, including muscle-induced wrinkles, glabellar lines, and realistic mouth cavity geometry, with high fidelity. Our method achieves state-of-the-art performance under extreme pose and expression variations, demonstrating strong generalization in challenging head reenactment settings. --- ### UniRec-0.1B: Unified Text and Formula Recognition with 0.1B Parameters Authors: Yongkun Du, Zhineng Chen, Yazhen Xie, Weikang Bai, Hao Feng, Wei Shi, Yuchen Su, Can Huang, Yu-Gang Jiang Published: 2025-12-24 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.21095v2 Text and formulas constitute the core informational components of many documents. Accurately and efficiently recognizing both is crucial for developing robust and generalizable document parsing systems. Recently, vision-language models (VLMs) have achieved impressive unified recognition of text and formulas. However, they are large-sized and computationally demanding, restricting their usage in many applications. In this paper, we propose UniRec-0.1B, a unified recognition model with only 0.1B parameters. It is capable of performing text and formula recognition at multiple levels, including characters, words, lines, paragraphs, and documents. To implement this task, we first establish UniRec40M, a large-scale dataset comprises 40 million text, formula and mixed samples, enabling the training of a powerful yet lightweight model. Secondly, we identify two challenges when building such a lightweight but unified expert model. They are: structural variability across levels and semantic entanglement between textual and formulaic content. To tackle these, we introduce a hierarchical supervision training that explicitly guides structural comprehension, and a semantic-decoupled tokenizer that separates text and formula representations. Finally, we develop a comprehensive evaluation benchmark covering Chinese and English documents from multiple domains and with multiple levels. Experimental results on this and public benchmarks demonstrate that UniRec-0.1B outperforms both general-purpose VLMs and leading document parsing expert models, while achieving 2-9x speedup, validating its effectiveness and efficiency. Codebase and Dataset: https://github.com/Topdu/OpenOCR. --- ### UniRec-0.1B: Unified Text and Formula Recognition with 0.1B Parameters Authors: Yongkun Du, Zhineng Chen, Yazhen Xie, Weikang Baiand Hao Feng, Wei Shi, Yuchen Su, Can Huang, Yu-Gang Jiang Published: 2025-12-24 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.21095v1 Text and formulas constitute the core informational components of many documents. Accurately and efficiently recognizing both is crucial for developing robust and generalizable document parsing systems. Recently, vision-language models (VLMs) have achieved impressive unified recognition of text and formulas. However, they are large-sized and computationally demanding, restricting their usage in many applications. In this paper, we propose UniRec-0.1B, a unified recognition model with only 0.1B parameters. It is capable of performing text and formula recognition at multiple levels, including characters, words, lines, paragraphs, and documents. To implement this task, we first establish UniRec40M, a large-scale dataset comprises 40 million text, formula and their mix samples, enabling the training of a powerful yet lightweight model. Secondly, we identify two challenges when building such a lightweight but unified expert model. They are: structural variability across hierarchies and semantic entanglement between textual and formulaic content. To tackle these, we introduce a hierarchical supervision training that explicitly guides structural comprehension, and a semantic-decoupled tokenizer that separates text and formula representations. Finally, we develop a comprehensive evaluation benchmark covering Chinese and English documents from multiple domains and with multiple levels. Experimental results on this and public benchmarks demonstrate that UniRec-0.1B outperforms both general-purpose VLMs and leading document parsing expert models, while achieving a 2-9$\times$ speedup, validating its effectiveness and efficiency. Codebase and Dataset: https://github.com/Topdu/OpenOCR. --- ### T2AV-Compass: Towards Unified Evaluation for Text-to-Audio-Video Generation Authors: Zhe Cao, Tao Wang, Jiaming Wang, Yanghai Wang, Yuanxing Zhang, Jiahao Wang, Jialu Chen, Miao Deng, Yubin Guo, Chenxi Liao, Yize Zhang, Zhaoxiang Zhang et al. Published: 2025-12-24 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.21094v3 Text-to-Audio-Video (T2AV) generation aims to synthesize temporally coherent video and semantically synchronized audio from natural language, yet its evaluation remains fragmented, often relying on unimodal metrics or narrowly scoped benchmarks that fail to capture cross-modal alignment, instruction following, and perceptual realism under complex prompts. To address this limitation, we present T2AV-Compass, a unified benchmark for comprehensive evaluation of T2AV systems, consisting of 500 diverse and complex prompts constructed via a taxonomy-driven pipeline to ensure semantic richness and physical plausibility. Besides, T2AV-Compass introduces a dual-level evaluation framework that integrates objective signal-level metrics for video quality, audio quality, and cross-modal alignment with a subjective MLLM-as-a-Judge protocol for instruction following and realism assessment. Extensive evaluation of 11 representative T2AVsystems reveals that even the strongest models fall substantially short of human-level realism and cross-modal consistency, with persistent failures in audio realism, fine-grained synchronization, instruction following, etc. These results indicate significant improvement room for future models and highlight the value of T2AV-Compass as a challenging and diagnostic testbed for advancing text-to-audio-video generation. --- ### T2AV-Compass: Towards Unified Evaluation for Text-to-Audio-Video Generation Authors: Zhe Cao, Tao Wang, Jiaming Wang, Yanghai Wang, Yuanxing Zhang, Jialu Chen, Miao Deng, Jiahao Wang, Yubin Guo, Chenxi Liao, Yize Zhang, Zhaoxiang Zhang et al. Published: 2025-12-24 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.21094v1 Text-to-Audio-Video (T2AV) generation aims to synthesize temporally coherent video and semantically synchronized audio from natural language, yet its evaluation remains fragmented, often relying on unimodal metrics or narrowly scoped benchmarks that fail to capture cross-modal alignment, instruction following, and perceptual realism under complex prompts. To address this limitation, we present T2AV-Compass, a unified benchmark for comprehensive evaluation of T2AV systems, consisting of 500 diverse and complex prompts constructed via a taxonomy-driven pipeline to ensure semantic richness and physical plausibility. Besides, T2AV-Compass introduces a dual-level evaluation framework that integrates objective signal-level metrics for video quality, audio quality, and cross-modal alignment with a subjective MLLM-as-a-Judge protocol for instruction following and realism assessment. Extensive evaluation of 11 representative T2AVsystems reveals that even the strongest models fall substantially short of human-level realism and cross-modal consistency, with persistent failures in audio realism, fine-grained synchronization, instruction following, etc. These results indicate significant improvement room for future models and highlight the value of T2AV-Compass as a challenging and diagnostic testbed for advancing text-to-audio-video generation. --- ### Hierarchical Modeling Approach to Fast and Accurate Table Recognition Authors: Takaya Kawakatsu Published: 2025-12-24 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2512.21083v1 The extraction and use of diverse knowledge from numerous documents is a pressing challenge in intelligent information retrieval. Documents contain elements that require different recognition methods. Table recognition typically consists of three subtasks, namely table structure, cell position and cell content recognition. Recent models have achieved excellent recognition with a combination of multi-task learning, local attention, and mutual learning. However, their effectiveness has not been fully explained, and they require a long period of time for inference. This paper presents a novel multi-task model that utilizes non-causal attention to capture the entire table structure, and a parallel inference algorithm for faster cell content inference. The superiority is demonstrated both visually and statistically on two large public datasets. --- ### Dyna-Style Reinforcement Learning Modeling and Control of Non-linear Dynamics Authors: Karim Abdelsalam, Zeyad Gamal, Ayman El-Badawy Published: 2025-12-24 Categories: eess.SY, cs.LG arXiv: https://arxiv.org/abs/2512.21081v1 Controlling systems with complex, nonlinear dynamics poses a significant challenge, particularly in achieving efficient and robust control. In this paper, we propose a Dyna-Style Reinforcement Learning control framework that integrates Sparse Identification of Nonlinear Dynamics (SINDy) with Twin Delayed Deep Deterministic Policy Gradient (TD3) reinforcement learning. SINDy is used to identify a data-driven model of the system, capturing its key dynamics without requiring an explicit physical model. This identified model is used to generate synthetic rollouts that are periodically injected into the reinforcement learning replay buffer during training on the real environment, enabling efficient policy learning with limited data available. By leveraging this hybrid approach, we mitigate the sample inefficiency of traditional model-free reinforcement learning methods while ensuring accurate control of nonlinear systems. To demonstrate the effectiveness of this framework, we apply it to a bi-rotor system as a case study, evaluating its performance in stabilization and trajectory tracking. The results show that our SINDy-TD3 approach achieves superior accuracy and robustness compared to direct reinforcement learning techniques, highlighting the potential of combining data-driven modeling with reinforcement learning for complex dynamical systems. --- ### LLM Personas as a Substitute for Field Experiments in Method Benchmarking Authors: Enoch Hyunwook Kang Published: 2025-12-24 Categories: cs.AI, cs.LG, econ.EM arXiv: https://arxiv.org/abs/2512.21080v3 Field experiments (A/B tests) are often the most credible benchmark for methods (algorithms) in societal systems, but their cost and latency bottleneck rapid methodological progress. LLM-based persona simulation offers a cheap synthetic alternative, yet it is unclear whether replacing humans with personas preserves the benchmark interface that adaptive methods optimize against. We prove an if-and-only-if characterization: when (i) methods observe only the aggregate outcome (aggregate-only observation) and (ii) evaluation depends only on the submitted artifact and not on the method's identity or provenance (method-blind evaluation), swapping humans for personas is just panel change from the method's point of view, indistinguishable from changing the evaluation population (e.g., New York to Jakarta). Furthermore, we move from validity to usefulness: we define an information-theoretic discriminability of the induced aggregate channel and show that making persona benchmarking as decision-relevant as a field experiment is fundamentally a sample-size question, yielding explicit bounds on the number of independent persona evaluations required to reliably distinguish meaningfully different methods at a chosen resolution. --- ### LLM Personas as a Substitute for Field Experiments in Method Benchmarking Authors: Enoch Hyunwook Kang Published: 2025-12-24 Categories: cs.AI, cs.LG, econ.EM arXiv: https://arxiv.org/abs/2512.21080v1 Field experiments (A/B tests) are often the most credible benchmark for methods in societal systems, but their cost and latency create a major bottleneck for iterative method development. LLM-based persona simulation offers a cheap synthetic alternative, yet it is unclear whether replacing humans with personas preserves the benchmark interface that adaptive methods optimize against. We prove an if-and-only-if characterization: when (i) methods observe only the aggregate outcome (aggregate-only observation) and (ii) evaluation depends only on the submitted artifact and not on the algorithm's identity or provenance (algorithm-blind evaluation), swapping humans for personas is just panel change from the method's point of view, indistinguishable from changing the evaluation population (e.g., New York to Jakarta). Furthermore, we move from validity to usefulness: we define an information-theoretic discriminability of the induced aggregate channel and show that making persona benchmarking as decision-relevant as a field experiment is fundamentally a sample-size question, yielding explicit bounds on the number of independent persona evaluations required to reliably distinguish meaningfully different methods at a chosen resolution. --- ### UniPR-3D: Towards Universal Visual Place Recognition with Visual Geometry Grounded Transformer Authors: Tianchen Deng, Xun Chen, Ziming Li, Hongming Shen, Shuhao Zhai, Danwei Wang, Javier Civera, Hesheng Wang Published: 2025-12-24 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.21078v3 Visual Place Recognition (VPR) has been traditionally formulated as a single-image retrieval task. Using multiple views offers clear advantages, yet this setting remains relatively underexplored and existing methods often struggle to generalize across diverse environments. In this work we introduce UniPR-3D, the first VPR architecture that effectively integrates information from multiple views. UniPR-3D builds on a VGGT backbone capable of encoding multi-view 3D representations, which we adapt by designing feature aggregators and fine-tune for the place recognition task. To construct our descriptor, we jointly leverage the 3D tokens and intermediate 2D tokens produced by VGGT. Based on their distinct characteristics, we design dedicated aggregation modules for 2D and 3D features, allowing our descriptor to capture fine-grained texture cues while also reasoning across viewpoints. To further enhance generalization, we incorporate both single- and multi-frame aggregation schemes, along with a variable-length sequence retrieval strategy. Our experiments show that UniPR-3D sets a new state of the art, outperforming both single- and multi-view baselines and highlighting the effectiveness of geometry-grounded tokens for VPR. Our code and models will be made publicly available on Github https://github.com/dtc111111/UniPR-3D. --- ### UniPR-3D: Towards Universal Visual Place Recognition with Visual Geometry Grounded Transformer Authors: Tianchen Deng, Xun Chen, Ziming Li, Hongming Shen, Danwei Wang, Javier Civera, Hesheng Wang Published: 2025-12-24 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.21078v1 Visual Place Recognition (VPR) has been traditionally formulated as a single-image retrieval task. Using multiple views offers clear advantages, yet this setting remains relatively underexplored and existing methods often struggle to generalize across diverse environments. In this work we introduce UniPR-3D, the first VPR architecture that effectively integrates information from multiple views. UniPR-3D builds on a VGGT backbone capable of encoding multi-view 3D representations, which we adapt by designing feature aggregators and fine-tune for the place recognition task. To construct our descriptor, we jointly leverage the 3D tokens and intermediate 2D tokens produced by VGGT. Based on their distinct characteristics, we design dedicated aggregation modules for 2D and 3D features, allowing our descriptor to capture fine-grained texture cues while also reasoning across viewpoints. To further enhance generalization, we incorporate both single- and multi-frame aggregation schemes, along with a variable-length sequence retrieval strategy. Our experiments show that UniPR-3D sets a new state of the art, outperforming both single- and multi-view baselines and highlighting the effectiveness of geometry-grounded tokens for VPR. Our code and models will be made publicly available on Github https://github.com/dtc111111/UniPR-3D. --- ### Hierarchical Book Organization for Learning-Resource Discovery using Dual-Path Graph Convolutions Authors: Suraj Kumar, Utsav Kumar Nareti, Soumi Chattopadhyay, Chandranath Adak, Prolay Mallick Published: 2025-12-24 Categories: cs.IR, cs.LG, cs.MM arXiv: https://arxiv.org/abs/2512.21076v2 The growing availability of books and textual materials in digital learning environments necessitates reliable semantic organization to support scalable resource management and discovery. However, existing book classification approaches typically formulate genre prediction as a flat classification problem, overlooking both the hierarchical organization of literary categories and the semantic discrepancy between authoritative book descriptions and subjective crowd-sourced reviews. We propose {\titleabbr}, a hierarchical book classification framework for structured learning-resource organization that reformulates genre prediction as coarse-to-fine semantic reasoning over heterogeneous textual evidence. HiGeMine first performs blurb-guided semantic refinement to preserve semantically consistent reviews while suppressing noisy and genre-irrelevant interpretations. It then performs semantic-role-separated hierarchical graph reasoning through independent propagation branches for blurbs and reviews, enabling separate modeling of narrative and interpretive semantics during hierarchical inference. A coarse-grained level-1 classifier first distinguishes fiction from non-fiction, followed by domain-specialized level-2 multi-label classifiers for fine-grained genre prediction. HiGeMine captures relationships among fine-grained resource categories through a structured label co-occurrence graph and genre-conditioned semantic representations. To facilitate systematic evaluation, we curate a new hierarchical multi-label Goodreads benchmark containing paired blurbs and reviews. Experiments against hierarchical classifiers, graph-based approaches, pretrained LMs, and LLMs demonstrate the effectiveness of HiGeMine for reliable hierarchical book classification, thereby providing a scalable foundation for organizing and discovering textual resources in digital learning environments. --- ### Blurb-Refined Inference from Crowdsourced Book Reviews using Hierarchical Genre Mining with Dual-Path Graph Convolutions Authors: Suraj Kumar, Utsav Kumar Nareti, Soumi Chattopadhyay, Chandranath Adak, Prolay Mallick Published: 2025-12-24 Categories: cs.IR, cs.LG, cs.MM arXiv: https://arxiv.org/abs/2512.21076v1 Accurate book genre classification is fundamental to digital library organization, content discovery, and personalized recommendation. Existing approaches typically model genre prediction as a flat, single-label task, ignoring hierarchical genre structure and relying heavily on noisy, subjective user reviews, which often degrade classification reliability. We propose HiGeMine, a two-phase hierarchical genre mining framework that robustly integrates user reviews with authoritative book blurbs. In the first phase, HiGeMine employs a zero-shot semantic alignment strategy to filter reviews, retaining only those semantically consistent with the corresponding blurb, thereby mitigating noise, bias, and irrelevance. In the second phase, we introduce a dual-path, two-level graph-based classification architecture: a coarse-grained Level-1 binary classifier distinguishes fiction from non-fiction, followed by Level-2 multi-label classifiers for fine-grained genre prediction. Inter-genre dependencies are explicitly modeled using a label co-occurrence graph, while contextual representations are derived from pretrained language models applied to the filtered textual content. To facilitate systematic evaluation, we curate a new hierarchical book genre dataset. Extensive experiments demonstrate that HiGeMine consistently outperformed strong baselines across hierarchical genre classification tasks. The proposed framework offers a principled and effective solution for leveraging both structured and unstructured textual data in hierarchical book genre analysis. --- ### Feature Learning Dynamics in Infinite-Depth Neural Networks Authors: Zihan Yao, Ruoyu Wu, Tianxiang Gao Published: 2025-12-24 Categories: cs.LG, cs.AI, math.PR, stat.ML arXiv: https://arxiv.org/abs/2512.21075v3 Deep neural networks have achieved remarkable success in practice, yet a mechanistic understanding of how features evolve during training remains incomplete, especially in the large-depth limit. For ResNets under depth-$μ$P scaling, prior work treats the layer index $\ell$ as a continuous time $t_\ell = \ell/L$, yielding SDE descriptions of the training dynamics. A key unresolved issue is that backpropagation reuses each forward weight matrix $W_\ell$ through its transpose $W_\ell^\top$, creating correlations between forward features and backward gradients whose behavior and role in feature learning remain unclear. We study this reused-weight forward--backward coupling in one-layer ResNets under depth-$μ$P. Using conditional Gaussian representations, we explicitly separate the coupling terms induced by weight reuse from decoupled Gaussian fluctuations before taking any network limit. At initialization, we prove that the coupling is a finite-width effect and vanishes at rate $O(n^{-1})$, uniformly over depth. During training, however, SGD induces a nontrivial forward--backward correlation term that survives the infinite-width limit. The key depth effect is that, under depth-$μ$P scaling, this surviving term is higher order in depth and its accumulated contribution over layers becomes negligible as $L\to\infty$. This depth-induced suppression motivates Neural Feature Dynamics (NFD), a forward--backward SDE system with decoupled backward weights that retains the feature-gradient covariance structure generated during training. Under nondegeneracy assumptions, we prove that the finite-network training dynamics converge to its NFD limit with an $O(L^{-1})$ depth-discretization error, while the reused-weight coupling term has a faster $O(L^{-2})$ decay. These results provide a rigorous infinite-depth limit for the feature-learning dynamics of one-layer ResNets under depth-$μ$P. --- ### Understanding Scaling Laws in Deep Neural Networks via Feature Learning Dynamics Authors: Zihan Yao, Ruoyu Wu, Tianxiang Gao Published: 2025-12-24 Categories: cs.LG, cs.AI, math.PR, stat.ML arXiv: https://arxiv.org/abs/2512.21075v1 The empirical success of deep learning is often attributed to scaling laws that predict consistent gains as model, data, and compute grow; however, large models can exhibit training instability and diminishing returns, suggesting that scaling laws describe what success looks like but not when and why scaling succeeds or fails. A central obstacle is the lack of a rigorous understanding of feature learning at large depth. While muP characterizes feature-learning dynamics in the infinite-width limit and enables hyperparameter transfer across width, its depth extension (depth-muP) breaks down for residual blocks with more than one internal layer. We derive Neural Feature Dynamics (NFD) for ResNets with single-layer residual blocks, characterizing feature learning via a coupled forward-backward stochastic system in the joint infinite-width and infinite-depth limit. In this regime, NFD identifies when scaling-law trends persist and explains diminishing returns. It also reveals a vanishing mechanism induced by the 1/sqrt(depth) residual scaling under which the gradient-independence assumption (GIA), known to fail during training at finite depth, becomes provably valid again at infinite depth, yielding an analytically tractable regime for end-to-end feature learning. Motivated by this insight, we study two-layer residual blocks and show that the same mechanism causes feature-learning collapse in the first internal layer at large depth, providing a structural explanation for the empirical failure of depth-muP. Based on this diagnosis, we propose a depth-aware learning-rate correction that counteracts the collapse and empirically restores depth-wise hyperparameter transfer, yielding stronger performance in deeper ResNets. --- ### Agentic Explainable Artificial Intelligence (Agentic XAI) Approach To Explore Better Explanation: A Case Study in Decision Support for Rice Cultivation in Japan Authors: Tomoaki Yamaguchi, Yutong Zhou, Masahiro Ryo, Keisuke Katsura Published: 2025-12-24 Categories: cs.AI, cs.HC arXiv: https://arxiv.org/abs/2512.21066v4 Explainable artificial intelligence (XAI) reveals how explanatory variables relate to a response variable, yet communicating XAI outputs to laypersons remains difficult, limiting trust in AI-based predictions. Large language models (LLMs) can translate technical explanations into accessible narratives, but iterative refinement of XAI explanations by an autonomous LLM agent remains unexplored. This study proposes an agentic XAI framework that combines SHapley Additive exPlanations (SHAP) with iterative refinement by a multimodal LLM and tests it as an agricultural recommendation system on rice yield data from 28 fields in Japan. From a SHAP result, the agent explored additional analyses across 11 refinement rounds (Rounds 0-10). Crop scientists (n = 12) and LLM judges (n = 14) scored every round on seven criteria: Specificity, Clarity, Conciseness, Practicality, Contextual Relevance, Cost Consideration, and Crop Science Credibility. Both groups found that refinement raised the average score by 30-33% over Round 0, peaking at Rounds 3-4, after which quality declined, below the starting point for crop scientists. Refinement therefore requires strategic early stopping, which challenges assumptions of monotonic improvement. Criterion-level trajectories indicate a bias-variance trade-off. Early rounds lacked Specificity (bias), whereas excessive iteration eroded Conciseness and raised Cost Consideration through ungrounded economic reasoning (variance). The LLM judges overscored every criterion by 1.4-2.3 points but largely preserved the experts' ranking of rounds (Spearman $ρ$ = 0.58-0.90), so screened LLM judges can flag the quality peak despite unreliable absolute scores. Trustworthy agentic XAI also needs expert-anchored screening of LLM judges and transparent, verifiable refinement records. ---