# Research Papers — 2025-12 · bundle page 142 of 175

> Full-text markdown of papers #7051–7100 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/142
[← Previous bundle](https://aipod.dev/papers/archive/2025-12/page/141.md)
[Next bundle →](https://aipod.dev/papers/archive/2025-12/page/143.md)

### MedDIFT: Multi-Scale Diffusion-Based Correspondence in 3D Medical Imaging

Authors: Xingyu Zhang, Anna Reithmeir, Fryderyk Kögl, Rickmer Braren, Julia A. Schnabel, Daniel M. Lang
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.05571v2

Accurate spatial correspondence between medical images is essential for longitudinal analysis, lesion tracking, and image-guided interventions. Medical image registration methods rely on local intensity-based similarity measures, which fail to capture global semantic structure and often yield mismatches in low-contrast or anatomically variable regions. Recent advances in diffusion models suggest that their intermediate representations encode rich geometric and semantic information. We present MedDIFT, a training-free 3D correspondence framework that leverages multi-scale features from a pretrained latent medical diffusion model as voxel descriptors. MedDIFT fuses diffusion activations into rich voxel-wise descriptors and matches them via cosine similarity, with an optional local-search prior. On a publicly available lung CT dataset, MedDIFT shows promising capability in identifying anatomical correspondence without requiring any task-specific model training. Ablation experiments confirm that multi-level feature fusion and modest diffusion noise improve performance. Code is available online.

---

### Wasserstein distance based semi-supervised manifold learning and application to GNSS multi-path detection

Authors: Antoine Blais, Nicolas Couëllan
Published: 2025-12-05
Categories: cs.LG, eess.SP, stat.ML
arXiv: https://arxiv.org/abs/2512.05567v1

The main objective of this study is to propose an optimal transport based semi-supervised approach to learn from scarce labelled image data using deep convolutional networks. The principle lies in implicit graph-based transductive semi-supervised learning where the similarity metric between image samples is the Wasserstein distance. This metric is used in the label propagation mechanism during learning. We apply and demonstrate the effectiveness of the method on a GNSS real life application. More specifically, we address the problem of multi-path interference detection. Experiments are conducted under various signal conditions. The results show that for specific choices of hyperparameters controlling the amount of semi-supervision and the level of sensitivity to the metric, the classification accuracy can be significantly improved over the fully supervised training method.

---

### ProPhy: Progressive Physical Alignment for Dynamic World Simulation

Authors: Zijun Wang, Panwen Hu, Jing Wang, Terry Jingchen Zhang, Yuhao Cheng, Long Chen, Yiqiang Yan, Zutao Jiang, Hanhui Li, Xiaodan Liang
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.05564v2

Recent advances in video generation have shown remarkable potential for constructing world simulators. However, current models still struggle to produce physically consistent results, particularly when handling large-scale or complex dynamics. This limitation arises primarily because existing approaches respond isotropically to physical prompts and neglect the fine-grained alignment between generated content and localized physical cues. To address these challenges, we propose ProPhy, a Progressive Physical Alignment Framework that enables explicit physics-aware conditioning and anisotropic generation. ProPhy employs a two-stage Mixture-of-Physics-Experts mechanism for discriminative physical prior extraction, where Semantic Experts infer semantic-level physical principles from textual descriptions, and Refinement Experts capture token-level physical dynamics. This mechanism allows the model to learn fine-grained, physics-aware video representations that better reflect underlying physical laws. Furthermore, we introduce a physical alignment strategy that transfers the physical reasoning capabilities of vision-language models into the Refinement Experts, facilitating a more accurate representation of dynamic physical phenomena. Extensive experiments on physics-aware video generation benchmarks demonstrate that ProPhy produces more realistic, dynamic, and physically coherent results than existing state-of-the-art methods.

---

### 2K-Characters-10K-Stories: A Quality-Gated Stylized Narrative Dataset with Disentangled Control and Sequence Consistency

Authors: Xingxi Yin, Yicheng Li, Gong Yan, Chenglin Li, Jian Zhao, Cong Huang, Yue Deng, Yin Zhang
Published: 2025-12-05
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2512.05557v1

Sequential identity consistency under precise transient attribute control remains a long-standing challenge in controllable visual storytelling. Existing datasets lack sufficient fidelity and fail to disentangle stable identities from transient attributes, limiting structured control over pose, expression, and scene composition and thus constraining reliable sequential synthesis. To address this gap, we introduce \textbf{2K-Characters-10K-Stories}, a multi-modal stylized narrative dataset of \textbf{2{,}000} uniquely stylized characters appearing across \textbf{10{,}000} illustration stories. It is the first dataset that pairs large-scale unique identities with explicit, decoupled control signals for sequential identity consistency. We introduce a \textbf{Human-in-the-Loop pipeline (HiL)} that leverages expert-verified character templates and LLM-guided narrative planning to generate highly-aligned structured data. A \textbf{decoupled control} scheme separates persistent identity from transient attributes -- pose and expression -- while a \textbf{Quality-Gated loop} integrating MMLM evaluation, Auto-Prompt Tuning, and Local Image Editing enforces pixel-level consistency. Extensive experiments demonstrate that models fine-tuned on our dataset achieves performance comparable to closed-source models in generating visual narratives.

---

### Beyond Linear Surrogates: High-Fidelity Local Explanations for Black-Box Models

Authors: Sanjeev Shrestha, Rahul Dubey, Hui Liu
Published: 2025-12-05
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.05556v3

With the increasing complexity of black-box machine learning models and their adoption in high-stakes areas, it is critical to provide explanations for their predictions. Existing local explanation methods lack in generating high-fidelity explanations. This paper proposes a novel local model agnostic explanation method to generate high-fidelity explanations using multivariate adaptive regression splines (MARS) and N-ball sampling strategies. MARS is used to model non-linear local boundaries that effectively captures the underlying behavior of the reference model, thereby enhancing the local fidelity. The N-ball sampling technique samples perturbed samples directly from a desired distribution instead of reweighting, leading to further improvement in the faithfulness. The performance of the proposed method was computed in terms of root mean squared error (RMSE) and evaluated on five different benchmark datasets with different kernel width. Experimental results show that the proposed method achieves higher local surrogate fidelity compared to baseline local explanation methods, with an average reduction of 32% in root mean square error, indicating more accurate local approximations of the black-box model. Additionally, statistical analysis shows that across all benchmark datasets, the proposed approach results were statistically significantly better. This paper advances the field of explainable AI by providing insights that can benefit the broader research and practitioner community.

---

### Conscious Gaze: Adaptive Attention Mechanisms for Hallucination Mitigation in Vision-Language Models

Authors: Weijue Bu, Guan Yuan, Guixian Zhang
Published: 2025-12-05
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2512.05546v1

Large Vision-Language Models (VLMs) often exhibit text inertia, where attention drifts from visual evidence toward linguistic priors, resulting in object hallucinations. Existing decoding strategies intervene only at the output logits and thus cannot correct internal reasoning drift, while recent internal-control methods based on heuristic head suppression or global steering vectors lack principled grounding. We introduce Conscious Gaze (CG-VLM), a training-free, inference-time framework that converts game-theoretic interpretability into actionable decoding control. A Cognitive Demand Sensor built on Harsanyi interactions estimates instantaneous vision-text synergy and identifies moments when visual grounding is necessary. Conditioned on this signal, a Focused Consensus Induction module selectively reorients mid-layer attention toward visual tokens before collapse into text priors. CG-VLM achieves state-of-the-art results on POPE and CHAIR across InstructBLIP, LLaVA, Qwen-VL, and mPLUG, while preserving general capabilities, demonstrating that token-level sensing enables precise, context-aware intervention without compromising foundational knowledge.

---

### RoBoN: Routed Online Best-of-n for Test-Time Scaling with Multiple LLMs

Authors: Jonathan Geuter, Gregor Kornhardt
Published: 2025-12-05
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.05542v1

Best-of-$n$ is a widely used test-time scaling approach for LLM inference. Yet despite evidence that LLMs exhibit complementary strengths across tasks, traditionally best-of-$n$ relies on a single model to generate responses. We propose RoBoN (Routed Online Best-of-$n$), a sequential multi-LLM alternative to the prevailing single-model best-of-$n$. Given a suite of models $\{m_i\}_{i=1}^M$, RoBoN sequentially routes generations one-by-one across models, based on scores computed using a reward model and an agreement signal on the predicted responses. This online routing requires no additional training, keeps compute parity, and works with any plug-in reward model. Across reasoning benchmarks (MATH500, OlympiadBench, MinervaMath, GSM8K, MMLU), RoBoN consistently outperforms standard best-of-$n$ applied to each individual model for larger $n$, with gains of up to 3.4\% in absolute accuracy, and also improves over a uniform multi-model portfolio baseline. Our results indicate that diversity across models can be exploited at inference to improve best-of-$n$ performance over any constituent model alone, providing a simple, training-free path to test-time scaling with multiple LLMs.

---

### SCoNE: Spherical Consistent Neighborhoods Ensemble for Effective and Efficient Multi-View Anomaly Detection

Authors: Yang Xu, Hang Zhang, Yixiao Ma, Ye Zhu, Kai Ming Ting
Published: 2025-12-05
Categories: cs.LG
arXiv: https://arxiv.org/abs/2512.05540v1

The core problem in multi-view anomaly detection is to represent local neighborhoods of normal instances consistently across all views. Recent approaches consider a representation of local neighborhood in each view independently, and then capture the consistent neighbors across all views via a learning process. They suffer from two key issues. First, there is no guarantee that they can capture consistent neighbors well, especially when the same neighbors are in regions of varied densities in different views, resulting in inferior detection accuracy. Second, the learning process has a high computational cost of $\mathcal{O}(N^2)$, rendering them inapplicable for large datasets. To address these issues, we propose a novel method termed \textbf{S}pherical \textbf{C}onsistent \textbf{N}eighborhoods \textbf{E}nsemble (SCoNE). It has two unique features: (a) the consistent neighborhoods are represented with multi-view instances directly, requiring no intermediate representations as used in existing approaches; and (b) the neighborhoods have data-dependent properties, which lead to large neighborhoods in sparse regions and small neighborhoods in dense regions. The data-dependent properties enable local neighborhoods in different views to be represented well as consistent neighborhoods, without learning. This leads to $\mathcal{O}(N)$ time complexity. Empirical evaluations show that SCoNE has superior detection accuracy and runs orders-of-magnitude faster in large datasets than existing approaches.

---

### Ideal Observer for Segmentation of Dead Leaves Images

Authors: Swantje Mahncke, Malte Ott, Lars C. Reining, Thomas S. A. Wallis
Published: 2025-12-05
Categories: cs.CV, math.ST, stat.ME
arXiv: https://arxiv.org/abs/2512.05539v3

The visible parts of a scene are determined by occlusion among overlapping surfaces. Here we consider "dead leaves" models, which replicate this by independently sampling objects ("leaves") with position, shape, color, and texture and layering them until the image is covered. Building on prior theory, we present a self-contained framework that rigorously defines the dead leaves model and derives an analytical Bayesian ideal observer for partitioning finite pixel sets. The longest part of the paper spans the derivation of the prior probability, which elevates the observer beyond pixel-similarity methods by incorporating geometric information. These computations are practical only for small pixel sets (up to 9-10 pixels). We emphasize accessibility through step-by-step derivations, extensive visualizations, and examples. We empirically evaluate three tractable observers (prior-only, likelihood-only, and the full ideal observer), plus a random baseline on 108 dead leaves image datasets varying in texture intensity, leaf size, and image size. Likelihood-only performance falls with increasing texture intensity and image size. Prior-only performance falls with decreasing leaf size and increasing maximal image dimension. All model-based observers strongly outperform the random baseline, and the ideal observer consistently outperforms the others by combining both information sources. The model provides a principled upper bound on segmentation performance for limited pixel sets, enabling comparisons with human observers and algorithms.

---

### Automated Identification of Incidentalomas Requiring Follow-Up: A Multi-Anatomy Evaluation of LLM-Based and Supervised Approaches

Authors: Namu Park, Farzad Ahmed, Zhaoyi Sun, Kevin Lybarger, Ethan Breinhorst, Julie Hu, Ozlem Uzuner, Martin Gunn, Meliha Yetisgen
Published: 2025-12-05
Categories: cs.CL
arXiv: https://arxiv.org/abs/2512.05537v1

Objective: To evaluate large language models (LLMs) against supervised baselines for fine-grained, lesion-level detection of incidentalomas requiring follow-up, addressing the limitations of current document-level classification systems. Methods: We utilized a dataset of 400 annotated radiology reports containing 1,623 verified lesion findings. We compared three supervised transformer-based encoders (BioClinicalModernBERT, ModernBERT, Clinical Longformer) against four generative LLM configurations (Llama 3.1-8B, GPT-4o, GPT-OSS-20b). We introduced a novel inference strategy using lesion-tagged inputs and anatomy-aware prompting to ground model reasoning. Performance was evaluated using class-specific F1-scores. Results: The anatomy-informed GPT-OSS-20b model achieved the highest performance, yielding an incidentaloma-positive macro-F1 of 0.79. This surpassed all supervised baselines (maximum macro-F1: 0.70) and closely matched the inter-annotator agreement of 0.76. Explicit anatomical grounding yielded statistically significant performance gains across GPT-based models (p < 0.05), while a majority-vote ensemble of the top systems further improved the macro-F1 to 0.90. Error analysis revealed that anatomy-aware LLMs demonstrated superior contextual reasoning in distinguishing actionable findings from benign lesions. Conclusion: Generative LLMs, when enhanced with structured lesion tagging and anatomical context, significantly outperform traditional supervised encoders and achieve performance comparable to human experts. This approach offers a reliable, interpretable pathway for automated incidental finding surveillance in radiology workflows.

---

### A Unified Theory of Sparse Dictionary Learning in Mechanistic Interpretability: Piecewise Biconvexity and Spurious Minima

Authors: Yiming Tang, Harshvardhan Saini, Zhaoqian Yao, Zheng Lin, Yizhen Liao, Jingyi Cui, Yisen Wang, Mengnan Du, Dianbo Liu
Published: 2025-12-05
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.05534v6

As AI models achieve remarkable capabilities across diverse domains, understanding what representations they learn and how they encode concepts has become increasingly important for both scientific progress and trustworthy deployment. Recent works in mechanistic interpretability have widely reported that neural networks represent meaningful concepts as linear directions in their representation spaces and often encode diverse concepts in superposition. Various sparse dictionary learning (SDL) methods, including sparse autoencoders, transcoders, and crosscoders, are utilized to address this by training auxiliary models with sparsity constraints to disentangle these superposed concepts into monosemantic features. These methods are the backbone of modern mechanistic interpretability, yet in practice they consistently produce polysemantic features, feature absorption, and dead neurons, with very limited theoretical understanding of why these phenomena occur. Existing theoretical work is limited to tied-weight sparse autoencoders, leaving the broader family of SDL methods without formal grounding. We develop the first unified theoretical framework that casts all major SDL variants as a single piecewise biconvex optimization problem, and characterize its global solution set, non-identifiability, and spurious optima. This analysis yields principled explanations for feature absorption and dead neurons. To expose these pathologies under full ground-truth access, we introduce the Linear Representation Bench. Guided by our theory, we propose feature anchoring, a novel technique that restores SDL identifiability, substantially improving feature recovery across synthetic benchmarks and real neural representations.

---

### IDK-S: Incremental Distributional Kernel for Streaming Anomaly Detection

Authors: Yang Xu, Yixiao Ma, Kaifeng Zhang, Zuliang Yang, Kai Ming Ting
Published: 2025-12-05
Categories: cs.LG
arXiv: https://arxiv.org/abs/2512.05531v1

Anomaly detection on data streams presents significant challenges, requiring methods to maintain high detection accuracy among evolving distributions while ensuring real-time efficiency. Here we introduce $\mathcal{IDK}$-$\mathcal{S}$, a novel $\mathbf{I}$ncremental $\mathbf{D}$istributional $\mathbf{K}$ernel for $\mathbf{S}$treaming anomaly detection that effectively addresses these challenges by creating a new dynamic representation in the kernel mean embedding framework. The superiority of $\mathcal{IDK}$-$\mathcal{S}$ is attributed to two key innovations. First, it inherits the strengths of the Isolation Distributional Kernel, an offline detector that has demonstrated significant performance advantages over foundational methods like Isolation Forest and Local Outlier Factor due to the use of a data-dependent kernel. Second, it adopts a lightweight incremental update mechanism that significantly reduces computational overhead compared to the naive baseline strategy of performing a full model retraining. This is achieved without compromising detection accuracy, a claim supported by its statistical equivalence to the full retrained model. Our extensive experiments on thirteen benchmarks demonstrate that $\mathcal{IDK}$-$\mathcal{S}$ achieves superior detection accuracy while operating substantially faster, in many cases by an order of magnitude, than existing state-of-the-art methods.

---

### MIND: Multi-rationale INtegrated Discriminative Reasoning Framework for Multi-modal Large Models

Authors: Chuang Yu, Jinmiao Zhao, Mingxuan Zhao, Yunpeng Liu, Xiujun Shu, Yuanhao Feng, Bo Wang, Xiangyu Yue
Published: 2025-12-05
Categories: cs.AI
arXiv: https://arxiv.org/abs/2512.05530v2

Recently, multimodal large language models (MLLMs) have been widely applied to reasoning tasks. However, they suffer from limited multi-rationale semantic modeling, insufficient logical robustness, and susceptibility to misleading cues. Therefore, we propose a Multi-rationale INtegrated Discriminative (MIND) reasoning framework, which is designed to endow MLLMs with human-like cognitive abilities of "Understand -> Rethink -> Correct", and achieves a paradigm evolution from passive imitation-based reasoning to active discriminative reasoning. Specifically, we introduce a Rationale Augmentation and Discrimination (RAD) paradigm, which provides a unified and extensible data foundation. Meanwhile, we design a Progressive Two-stage Correction Learning (P2CL) strategy. The first phase enhances multi-rationale positive learning, while the second phase enables active logic discrimination and correction. In addition, to mitigate representation entanglement in the multi-rationale semantic space, we propose a Multi-rationale Contrastive Alignment (MCA) optimization strategy. Extensive experiments show that our MIND achieves SOTA performance on multiple public datasets. Our data and code are available at https://github.com/YuChuang1205/MIND

---

### See in Depth: Training-Free Surgical Scene Segmentation with Monocular Depth Priors

Authors: Kunyi Yang, Qingyu Wang, Cheng Yuan, Yutong Ban
Published: 2025-12-05
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2512.05529v1

Pixel-wise segmentation of laparoscopic scenes is essential for computer-assisted surgery but difficult to scale due to the high cost of dense annotations. We propose depth-guided surgical scene segmentation (DepSeg), a training-free framework that utilizes monocular depth as a geometric prior together with pretrained vision foundation models. DepSeg first estimates a relative depth map with a pretrained monocular depth estimation network and proposes depth-guided point prompts, which SAM2 converts into class-agnostic masks. Each mask is then described by a pooled pretrained visual feature and classified via template matching against a template bank built from annotated frames. On the CholecSeg8k dataset, DepSeg improves over a direct SAM2 auto segmentation baseline (35.9% vs. 14.7% mIoU) and maintains competitive performance even when using only 10--20% of the object templates. These results show that depth-guided prompting and template-based classification offer an annotation-efficient segmentation approach.

---

### Decoding Selective Auditory Attention to Musical Elements in Ecologically Valid Music Listening

Authors: Taketo Akama, Zhuohao Zhang, Tsukasa Nagashima, Takagi Yutaka, Shun Minamikawa, Natalia Polouliakh
Published: 2025-12-05
Categories: q-bio.NC, cs.LG, cs.SD, eess.AS, eess.SP
arXiv: https://arxiv.org/abs/2512.05528v1

Art has long played a profound role in shaping human emotion, cognition, and behavior. While visual arts such as painting and architecture have been studied through eye tracking, revealing distinct gaze patterns between experts and novices, analogous methods for auditory art forms remain underdeveloped. Music, despite being a pervasive component of modern life and culture, still lacks objective tools to quantify listeners' attention and perceptual focus during natural listening experiences. To our knowledge, this is the first attempt to decode selective attention to musical elements using naturalistic, studio-produced songs and a lightweight consumer-grade EEG device with only four electrodes. By analyzing neural responses during real world like music listening, we test whether decoding is feasible under conditions that minimize participant burden and preserve the authenticity of the musical experience. Our contributions are fourfold: (i) decoding music attention in real studio-produced songs, (ii) demonstrating feasibility with a four-channel consumer EEG, (iii) providing insights for music attention decoding, and (iv) demonstrating improved model ability over prior work. Our findings suggest that musical attention can be decoded not only for novel songs but also across new subjects, showing performance improvements compared to existing approaches under our tested conditions. These findings show that consumer-grade devices can reliably capture signals, and that neural decoding in music could be feasible in real-world settings. This paves the way for applications in education, personalized music technologies, and therapeutic interventions.

---

### Credal and Interval Deep Evidential Classifications

Authors: Michele Caprio, Shireen K. Manchingal, Fabio Cuzzolin
Published: 2025-12-05
Categories: cs.LG, math.ST, stat.ME
arXiv: https://arxiv.org/abs/2512.05526v1

Uncertainty Quantification (UQ) presents a pivotal challenge in the field of Artificial Intelligence (AI), profoundly impacting decision-making, risk assessment and model reliability. In this paper, we introduce Credal and Interval Deep Evidential Classifications (CDEC and IDEC, respectively) as novel approaches to address UQ in classification tasks. CDEC and IDEC leverage a credal set (closed and convex set of probabilities) and an interval of evidential predictive distributions, respectively, allowing us to avoid overfitting to the training data and to systematically assess both epistemic (reducible) and aleatoric (irreducible) uncertainties. When those surpass acceptable thresholds, CDEC and IDEC have the capability to abstain from classification and flag an excess of epistemic or aleatoric uncertainty, as relevant. Conversely, within acceptable uncertainty bounds, CDEC and IDEC provide a collection of labels with robust probabilistic guarantees. CDEC and IDEC are trained using standard backpropagation and a loss function that draws from the theory of evidence. They overcome the shortcomings of previous efforts, and extend the current evidential deep learning literature. Through extensive experiments on MNIST, CIFAR-10 and CIFAR-100, together with their natural OoD shifts (F-MNIST/K-MNIST, SVHN/Intel, TinyImageNet), we show that CDEC and IDEC achieve competitive predictive accuracy, state-of-the-art OoD detection under epistemic and total uncertainty, and tight, well-calibrated prediction regions that expand reliably under distribution shift. An ablation over ensemble size further demonstrates that CDEC attains stable uncertainty estimates with only a small ensemble.

---

### Poodle: Seamlessly Scaling Down Large Language Models with Just-in-Time Model Replacement

Authors: Nils Strassenburg, Boris Glavic, Tilmann Rabl
Published: 2025-12-05
Categories: cs.DB, cs.LG
arXiv: https://arxiv.org/abs/2512.05525v3

Businesses increasingly rely on large language models (LLMs) to automate simple repetitive tasks instead of developing custom machine learning models. LLMs require few, if any, training examples and can be utilized by users without expertise in model development. However, this comes at the cost of substantially higher resource and energy consumption compared to smaller models, which often achieve similar predictive performance for simple tasks. In this paper, we present our vision for just-in-time model replacement (JITR), where, upon identifying a recurring task in calls to an LLM, the model is replaced transparently with a cheaper alternative that performs well for this specific task. JITR retains the ease of use and low development effort of LLMs, while saving significant cost and energy. We discuss the main challenges in realizing our vision regarding the identification of recurring tasks and the creation of a custom model. Specifically, we argue that model search and transfer learning will play a crucial role in JITR to efficiently identify and fine-tune models for a recurring task. Using our JITR prototype Poodle, we reduce inference time by up to 7.5x compared to a self- hosted LLM and save more than $2,200 per 1M requests compared to a flagship hosted LLM, while achieving accuracy competitive with the LLM baseline.

---

### VOST-SGG: VLM-Aided One-Stage Spatio-Temporal Scene Graph Generation

Authors: Chinthani Sugandhika, Chen Li, Deepu Rajan, Basura Fernando
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.05524v2

Spatio-temporal scene graph generation (ST-SGG) aims to model objects and their evolving relationships across video frames, enabling interpretable representations for downstream reasoning tasks such as video captioning and visual question answering. Despite recent advancements in DETR-style single-stage ST-SGG models, they still suffer from several key limitations. First, while these models rely on attention-based learnable queries as a core component, these learnable queries are semantically uninformed and instance-agnostically initialized. Second, these models rely exclusively on unimodal visual features for predicate classification. To address these challenges, we propose VOST-SGG, a VLM-aided one-stage ST-SGG framework that integrates the common sense reasoning capabilities of vision-language models (VLMs) into the ST-SGG pipeline. First, we introduce the dual-source query initialization strategy that disentangles what to attend to from where to attend, enabling semantically grounded what-where reasoning. Furthermore, we propose a multi-modal feature bank that fuses visual, textual, and spatial cues derived from VLMs for improved predicate classification. Extensive experiments on the Action Genome dataset demonstrate that our approach achieves state-of-the-art performance, validating the effectiveness of integrating VLM-aided semantic priors and multi-modal features for ST-SGG. We will release the code at https://github.com/LUNAProject22/VOST.

---

### User Negotiations of Authenticity, Ownership, and Governance on AI-Generated Video Platforms: Evidence from Sora

Authors: Bohui Shen, Shrikar Bhatta, Alex Ireebanije, Zexuan Liu, Abhinav Choudhry, Ece Gumusel, Kyrie Zhixuan Zhou
Published: 2025-12-05
Categories: cs.HC, cs.AI, cs.CY
arXiv: https://arxiv.org/abs/2512.05519v1

As AI-generated video platforms rapidly advance, ethical challenges such as copyright infringement emerge. This study examines how users make sense of AI-generated videos on OpenAI's Sora by conducting a qualitative content analysis of user comments. Through a thematic analysis, we identified four dynamics that characterize how users negotiate authenticity, authorship, and platform governance on Sora. First, users acted as critical evaluators of realism, assessing micro-details such as lighting, shadows, fluid motion, and physics to judge whether AI-generated scenes could plausibly exist. Second, users increasingly shifted from passive viewers to active creators, expressing curiosity about prompts, techniques, and creative processes. Text prompts were perceived as intellectual property, generating concerns about plagiarism and remixing norms. Third, users reported blurred boundaries between real and synthetic media, worried about misinformation, and even questioned the authenticity of other commenters, suspecting bot-generated engagement. Fourth, users contested platform governance: some perceived moderation as inconsistent or opaque, while others shared tactics for evading prompt censorship through misspellings, alternative phrasing, emojis, or other languages. Despite this, many users also enforced ethical norms by discouraging the misuse of real people's images or disrespectful content. Together, these patterns highlighted how AI-mediated platforms complicate notions of reality, creativity, and rule-making in emerging digital ecosystems. Based on the findings, we discuss governance challenges in Sora and how user negotiations inform future platform governance.

---

### Matching Ranks Over Probability Yields Truly Deep Safety Alignment

Authors: Jason Vega, Gagandeep Singh
Published: 2025-12-05
Categories: cs.CR, cs.AI
arXiv: https://arxiv.org/abs/2512.05518v2

Open-source Large Language Models (LLMs) play a critical role in the democratization of AI, yet their "open" nature introduces more avenues for malicious actors to misuse them for harmful purposes. A frustratingly easy but powerful technique known as the prefilling attack has been shown to effectively circumvent the safety alignment of frontier open-source LLMs by simply prefilling the assistant response with an affirmative prefix before decoding. A recent promising supervised fine-tuning defense proposes using a simple data augmentation scheme to achieve a "deep" safety alignment, allowing the model to generate natural language refusals immediately following harmful prefills. In this work, we show that a simple generalization of the prefilling attack, which we refer to as the Rank-Assisted Prefilling (RAP) attack, can effectively extract harmful content from models fine-tuned with the data augmentation defense by selecting low-probability "harmful" tokens from the top 20 predicted next tokens at each step (thus ignoring high-probability \enquote{refusal} tokens). We then propose a new perspective on achieving deep safety alignment by matching the token ranks in the underlying data augmentation target distribution (rather than just their probabilities), yielding a surprisingly simple approach to strengthening deep alignment we call PRefill attEntion STOpping (PRESTO) that regularizes the attention placed on harmful prefill tokens. Compared to just using data augmentation, PRESTO yields up to a 4.7x improvement in safety under RAP attacks across three popular LLMs. By achieving a stronger level of safety against practical and accessible attacks, our work paves a path towards safer open-source models. The source code accompanying this paper can be found at: https://github.com/uiuc-focal-lab/push-forward-alignment

---

### DashFusion: Dual-stream Alignment with Hierarchical Bottleneck Fusion for Multimodal Sentiment Analysis

Authors: Yuhua Wen, Qifei Li, Yingying Zhou, Yingming Gao, Zhengqi Wen, Jianhua Tao, Ya Li
Published: 2025-12-05
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2512.05515v1

Multimodal sentiment analysis (MSA) integrates various modalities, such as text, image, and audio, to provide a more comprehensive understanding of sentiment. However, effective MSA is challenged by alignment and fusion issues. Alignment requires synchronizing both temporal and semantic information across modalities, while fusion involves integrating these aligned features into a unified representation. Existing methods often address alignment or fusion in isolation, leading to limitations in performance and efficiency. To tackle these issues, we propose a novel framework called Dual-stream Alignment with Hierarchical Bottleneck Fusion (DashFusion). Firstly, dual-stream alignment module synchronizes multimodal features through temporal and semantic alignment. Temporal alignment employs cross-modal attention to establish frame-level correspondences among multimodal sequences. Semantic alignment ensures consistency across the feature space through contrastive learning. Secondly, supervised contrastive learning leverages label information to refine the modality features. Finally, hierarchical bottleneck fusion progressively integrates multimodal information through compressed bottleneck tokens, which achieves a balance between performance and computational efficiency. We evaluate DashFusion on three datasets: CMU-MOSI, CMU-MOSEI, and CH-SIMS. Experimental results demonstrate that DashFusion achieves state-of-the-art performance across various metrics, and ablation studies confirm the effectiveness of our alignment and fusion techniques. The codes for our experiments are available at https://github.com/ultramarineX/DashFusion.

---

### Know-Show: Benchmarking Video-Language Models on Spatio-Temporal Grounded Reasoning

Authors: Chinthani Sugandhika, Chen Li, Deepu Rajan, Basura Fernando
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.05513v3

Large Video-Language Models (Video-LMs) have achieved impressive progress in multimodal understanding, yet their reasoning remains weakly grounded in space and time. We present Know-Show, a new benchmark designed to evaluate spatio-temporal grounded reasoning, the ability of a model to reason about actions and their semantics while simultaneously grounding its inferences in visual and temporal evidence. Know-Show unifies reasoning and localization within a single evaluation framework consisting of five complementary scenarios across spatial (person, object, person-object, and hand-object) and temporal dimensions. Built from Charades, Action Genome, and Ego4D with 2.5K high-quality human-authored questions, the benchmark exposes significant gaps between current Video-LMs and human reasoning. To bridge this gap, we propose GRAM, a training-free plug-in that augments Video-LMs with fine-grained grounded reasoning through attention-based video token selection and explicit timestamp encoding. Extensive experiments across open and closed Video-LMs (e.g., Qwen, VideoR1, Gemini, and GPT-4o) reveal that existing models struggle to "show what they know" and vice versa. Know-Show establishes a unified standard for assessing grounded reasoning in video-language understanding and provides insights toward developing interpretable and reliable multimodal reasoning systems. We have released the dataset at https://github.com/LUNAProject22/Know-Show, and the code will be released in the same repository.

---

### Rethinking Infrared Small Target Detection: A Foundation-Driven Efficient Paradigm

Authors: Chuang Yu, Jinmiao Zhao, Yunpeng Liu, Yaokun Li, Xiujun Shu, Yuanhao Feng, Bo Wang, Yimian Dai, Xiangyu Yue
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.05511v1

While large-scale visual foundation models (VFMs) exhibit strong generalization across diverse visual domains, their potential for single-frame infrared small target (SIRST) detection remains largely unexplored. To fill this gap, we systematically introduce the frozen representations from VFMs into the SIRST task for the first time and propose a Foundation-Driven Efficient Paradigm (FDEP), which can seamlessly adapt to existing encoder-decoder-based methods and significantly improve accuracy without additional inference overhead. Specifically, a Semantic Alignment Modulation Fusion (SAMF) module is designed to achieve dynamic alignment and deep fusion of the global semantic priors from VFMs with task-specific features. Meanwhile, to avoid the inference time burden introduced by VFMs, we propose a Collaborative Optimization-based Implicit Self-Distillation (CO-ISD) strategy, which enables implicit semantic transfer between the main and lightweight branches through parameter sharing and synchronized backpropagation. In addition, to unify the fragmented evaluation system, we construct a Holistic SIRST Evaluation (HSE) metric that performs multi-threshold integral evaluation at both pixel-level confidence and target-level robustness, providing a stable and comprehensive basis for fair model comparison. Extensive experiments demonstrate that the SIRST detection networks equipped with our FDEP framework achieve state-of-the-art (SOTA) performance on multiple public datasets. Our code is available at https://github.com/YuChuang1205/FDEP-Framework

---

### Lyrics Matter: Exploiting the Power of Learnt Representations for Music Popularity Prediction

Authors: Yash Choudhary, Preeti Rao, Pushpak Bhattacharyya
Published: 2025-12-05
Categories: cs.SD, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2512.05508v1

Accurately predicting music popularity is a critical challenge in the music industry, offering benefits to artists, producers, and streaming platforms. Prior research has largely focused on audio features, social metadata, or model architectures. This work addresses the under-explored role of lyrics in predicting popularity. We present an automated pipeline that uses LLM to extract high-dimensional lyric embeddings, capturing semantic, syntactic, and sequential information. These features are integrated into HitMusicLyricNet, a multimodal architecture that combines audio, lyrics, and social metadata for popularity score prediction in the range 0-100. Our method outperforms existing baselines on the SpotGenTrack dataset, which contains over 100,000 tracks, achieving 9% and 20% improvements in MAE and MSE, respectively. Ablation confirms that gains arise from our LLM-driven lyrics feature pipeline (LyricsAENet), underscoring the value of dense lyric representations.

---

### GRASP: Graph Reasoning Agents for Systems Pharmacology with Human-in-the-Loop

Authors: Omid Bazgir, Vineeth Manthapuri, Ilia Rattsev, Mohammad Jafarnejad
Published: 2025-12-05
Categories: cs.LG
arXiv: https://arxiv.org/abs/2512.05502v1

Quantitative Systems Pharmacology (QSP) modeling is essential for drug development but it requires significant time investment that limits the throughput of domain experts. We present \textbf{GRASP} -- a multi-agent, graph-reasoning framework with a human-in-the-loop conversational interface -- that encodes QSP models as typed biological knowledge graphs and compiles them to executable MATLAB/SimBiology code while preserving units, mass balance, and physiological constraints. A two-phase workflow -- \textsc{Understanding} (graph reconstruction of legacy code) and \textsc{Action} (constraint-checked, language-driven modification) -- is orchestrated by a state machine with iterative validation. GRASP performs breadth-first parameter-alignment around new entities to surface dependent quantities and propose biologically plausible defaults, and it runs automatic execution/diagnostics until convergence. In head-to-head evaluations using LLM-as-judge, GRASP outperforms SME-guided CoT and ToT baselines across biological plausibility, mathematical correctness, structural fidelity, and code quality (\(\approx\)9--10/10 vs.\ 5--7/10). BFS alignment achieves F1 = 0.95 for dependency discovery, units, and range. These results demonstrate that graph-structured, agentic workflows can make QSP model development both accessible and rigorous, enabling domain experts to specify mechanisms in natural language without sacrificing biomedical fidelity.

---

### SEA-SafeguardBench: Evaluating AI Safety in SEA Languages and Cultures

Authors: Panuthep Tasawong, Jian Gang Ngui, Alham Fikri Aji, Trevor Cohn, Peerat Limkonchotiwat
Published: 2025-12-05
Categories: cs.CL
arXiv: https://arxiv.org/abs/2512.05501v1

Safeguard models help large language models (LLMs) detect and block harmful content, but most evaluations remain English-centric and overlook linguistic and cultural diversity. Existing multilingual safety benchmarks often rely on machine-translated English data, which fails to capture nuances in low-resource languages. Southeast Asian (SEA) languages are underrepresented despite the region's linguistic diversity and unique safety concerns, from culturally sensitive political speech to region-specific misinformation. Addressing these gaps requires benchmarks that are natively authored to reflect local norms and harm scenarios. We introduce SEA-SafeguardBench, the first human-verified safety benchmark for SEA, covering eight languages, 21,640 samples, across three subsets: general, in-the-wild, and content generation. The experimental results from our benchmark demonstrate that even state-of-the-art LLMs and guardrails are challenged by SEA cultural and harm scenarios and underperform when compared to English texts.

---

### SSDLabeler: Realistic semi-synthetic data generation for multi-label artifact classification in EEG

Authors: Taketo Akama, Akima Connelly, Shun Minamikawa, Natalia Polouliakh
Published: 2025-12-05
Categories: q-bio.NC, cs.LG, eess.SP
arXiv: https://arxiv.org/abs/2512.05500v1

EEG recordings are inherently contaminated by artifacts such as ocular, muscular, and environmental noise, which obscure neural activity and complicate preprocessing. Artifact classification offers advantages in stability and transparency, providing a viable alternative to ICA-based methods that enable flexible use alongside human inspections and across various applications. However, artifact classification is limited by its training data as it requires extensive manual labeling, which cannot fully cover the diversity of real-world EEG. Semi-synthetic data (SSD) methods have been proposed to address this limitation, but prior approaches typically injected single artifact types using ICA components or required separately recorded artifact signals, reducing both the realism of the generated data and the applicability of the method. To overcome these issues, we introduce SSDLabeler, a framework that generates realistic, annotated SSDs by decomposing real EEG with ICA, epoch-level artifact verification using RMS and PSD criteria, and reinjecting multiple artifact types into clean data. When applied to train a multi-label artifact classifier, it improved accuracy on raw EEG across diverse conditions compared to prior SSD and raw EEG training, establishing a scalable foundation for artifact handling that captures the co-occurrence and complexity of real EEG.

---

### Decoding with Structured Awareness: Integrating Directional, Frequency-Spatial, and Structural Attention for Medical Image Segmentation

Authors: Fan Zhang, Zhiwei Gu, Hua Wang
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.05494v1

To address the limitations of Transformer decoders in capturing edge details, recognizing local textures and modeling spatial continuity, this paper proposes a novel decoder framework specifically designed for medical image segmentation, comprising three core modules. First, the Adaptive Cross-Fusion Attention (ACFA) module integrates channel feature enhancement with spatial attention mechanisms and introduces learnable guidance in three directions (planar, horizontal, and vertical) to enhance responsiveness to key regions and structural orientations. Second, the Triple Feature Fusion Attention (TFFA) module fuses features from Spatial, Fourier and Wavelet domains, achieving joint frequency-spatial representation that strengthens global dependency and structural modeling while preserving local information such as edges and textures, making it particularly effective in complex and blurred boundary scenarios. Finally, the Structural-aware Multi-scale Masking Module (SMMM) optimizes the skip connections between encoder and decoder by leveraging multi-scale context and structural saliency filtering, effectively reducing feature redundancy and improving semantic interaction quality. Working synergistically, these modules not only address the shortcomings of traditional decoders but also significantly enhance performance in high-precision tasks such as tumor segmentation and organ boundary extraction, improving both segmentation accuracy and model generalization. Experimental results demonstrate that this framework provides an efficient and practical solution for medical image segmentation.

---

### WaterWave: Bridging Underwater Image Enhancement into Video Streams via Wavelet-based Temporal Consistency Field

Authors: Qi Zhu, Jingyi Zhang, Naishan Zheng, Wei Yu, Jinghao Zhang, Deyi Ji, Feng Zhao
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.05492v1

Underwater video pairs are fairly difficult to obtain due to the complex underwater imaging. In this case, most existing video underwater enhancement methods are performed by directly applying the single-image enhancement model frame by frame, but a natural issue is lacking temporal consistency. To relieve the problem, we rethink the temporal manifold inherent in natural videos and observe a temporal consistency prior in dynamic scenes from the local temporal frequency perspective. Building upon the specific prior and no paired-data condition, we propose an implicit representation manner for enhanced video signals, which is conducted in the wavelet-based temporal consistency field, WaterWave. Specifically, under the constraints of the prior, we progressively filter and attenuate the inconsistent components while preserving motion details and scenes, achieving a natural-flowing video. Furthermore, to represent temporal frequency bands more accurately, an underwater flow correction module is designed to rectify estimated flows considering the transmission in underwater scenes. Extensive experiments demonstrate that WaterWave significantly enhances the quality of videos generated using single-image underwater enhancements. Additionally, our method demonstrates high potential in downstream underwater tracking tasks, such as UOSTrack and MAT, outperforming the original video by a large margin, i.e., 19.7% and 9.7% on precise respectively.

---

### Turbulence Regression

Authors: Yingang Fan, Binjie Ding, Baiyi Chen
Published: 2025-12-05
Categories: cs.LG
arXiv: https://arxiv.org/abs/2512.05483v1

Air turbulence refers to the disordered and irregular motion state generated by drastic changes in velocity, pressure, or direction during airflow. Various complex factors lead to intricate low-altitude turbulence outcomes. Under current observational conditions, especially when using only wind profile radar data, traditional methods struggle to accurately predict turbulence states. Therefore, this paper introduces a NeuTucker decomposition model utilizing discretized data. Designed for continuous yet sparse three-dimensional wind field data, it constructs a low-rank Tucker decomposition model based on a Tucker neural network to capture the latent interactions within the three-dimensional wind field data. Therefore, two core ideas are proposed here: 1) Discretizing continuous input data to adapt to models like NeuTucF that require discrete data inputs. 2) Constructing a four-dimensional Tucker interaction tensor to represent all possible spatio-temporal interactions among different elevations and three-dimensional wind speeds. In estimating missing observations in real datasets, this discretized NeuTucF model demonstrates superior performance compared to various common regression models.

---

### Concept-based Explainable Data Mining with VLM for 3D Detection

Authors: Mai Tsujimoto
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.05482v1

Rare-object detection remains a challenging task in autonomous driving systems, particularly when relying solely on point cloud data. Although Vision-Language Models (VLMs) exhibit strong capabilities in image understanding, their potential to enhance 3D object detection through intelligent data mining has not been fully explored. This paper proposes a novel cross-modal framework that leverages 2D VLMs to identify and mine rare objects from driving scenes, thereby improving 3D object detection performance. Our approach synthesizes complementary techniques such as object detection, semantic feature extraction, dimensionality reduction, and multi-faceted outlier detection into a cohesive, explainable pipeline that systematically identifies rare but critical objects in driving scenes. By combining Isolation Forest and t-SNE-based outlier detection methods with concept-based filtering, the framework effectively identifies semantically meaningful rare objects. A key strength of this approach lies in its ability to extract and annotate targeted rare object concepts such as construction vehicles, motorcycles, and barriers. This substantially reduces the annotation burden and focuses only on the most valuable training samples. Experiments on the nuScenes dataset demonstrate that this concept-guided data mining strategy enhances the performance of 3D object detection models while utilizing only a fraction of the training data, with particularly notable improvements for challenging object categories such as trailers and bicycles compared with the same amount of random data. This finding has substantial implications for the efficient curation of datasets in safety-critical autonomous systems.

---

### UniFS: Unified Multi-Contrast MRI Reconstruction via Frequency-Spatial Fusion

Authors: Jialin Li, Yiwei Ren, Kai Pan, Dong Wei, Pujin Cheng, Xian Wu, Xiaoying Tang
Published: 2025-12-05
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2512.05481v1

Recently, Multi-Contrast MR Reconstruction (MCMR) has emerged as a hot research topic that leverages high-quality auxiliary modalities to reconstruct undersampled target modalities of interest. However, existing methods often struggle to generalize across different k-space undersampling patterns, requiring the training of a separate model for each specific pattern, which limits their practical applicability. To address this challenge, we propose UniFS, a Unified Frequency-Spatial Fusion model designed to handle multiple k-space undersampling patterns for MCMR tasks without any need for retraining. UniFS integrates three key modules: a Cross-Modal Frequency Fusion module, an Adaptive Mask-Based Prompt Learning module, and a Dual-Branch Complementary Refinement module. These modules work together to extract domain-invariant features from diverse k-space undersampling patterns while dynamically adapt to their own variations. Another limitation of existing MCMR methods is their tendency to focus solely on spatial information while neglect frequency characteristics, or extract only shallow frequency features, thus failing to fully leverage complementary cross-modal frequency information. To relieve this issue, UniFS introduces an adaptive prompt-guided frequency fusion module for k-space learning, significantly enhancing the model's generalization performance. We evaluate our model on the BraTS and HCP datasets with various k-space undersampling patterns and acceleration factors, including previously unseen patterns, to comprehensively assess UniFS's generalizability. Experimental results across multiple scenarios demonstrate that UniFS achieves state-of-the-art performance. Our code is available at https://github.com/LIKP0/UniFS.

---

### EmoStyle: Emotion-Driven Image Stylization

Authors: Jingyuan Yang, Zihuan Bai, Hui Huang
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.05478v1

Art has long been a profound medium for expressing emotions. While existing image stylization methods effectively transform visual appearance, they often overlook the emotional impact carried by styles. To bridge this gap, we introduce Affective Image Stylization (AIS), a task that applies artistic styles to evoke specific emotions while preserving content. We present EmoStyle, a framework designed to address key challenges in AIS, including the lack of training data and the emotion-style mapping. First, we construct EmoStyleSet, a content-emotion-stylized image triplet dataset derived from ArtEmis to support AIS. We then propose an Emotion-Content Reasoner that adaptively integrates emotional cues with content to learn coherent style queries. Given the discrete nature of artistic styles, we further develop a Style Quantizer that converts continuous style features into emotion-related codebook entries. Extensive qualitative and quantitative evaluations, including user studies, demonstrate that EmoStyle enhances emotional expressiveness while maintaining content consistency. Moreover, the learned emotion-aware style dictionary is adaptable to other generative tasks, highlighting its potential for broader applications. Our work establishes a foundation for emotion-driven image stylization, expanding the creative potential of AI-generated art.

---

### PERM EQ x GRAPH EQ: Equivariant Neural Networks for Quantum Molecular Learning

Authors: Saumya Biswas, Jiten Oswal
Published: 2025-12-05
Categories: cs.LG, cs.AI, quant-ph
arXiv: https://arxiv.org/abs/2512.05475v1

In hierarchal order of molecular geometry, we compare the performances of Geometric Quantum Machine Learning models. Two molecular datasets are considered: the simplistic linear shaped LiH-molecule and the trigonal pyramidal molecule NH3. Both accuracy and generalizability metrics are considered. A classical equivariant model is used as a baseline for the performance comparison. The comparative performance of Quantum Machine Learning models with no symmetry equivariance, rotational and permutational equivariance, and graph embedded permutational equivariance is investigated. The performance differentials and the molecular geometry in question reveals the criteria for choice of models for generalizability. Graph embedding of features is shown to be an effective pathway to greater trainability for geometric datasets. Permutational symmetric embedding is found to be the most generalizable quantum Machine Learning model for geometric learning.

---

### Unleashing Temporal Capacity of Spiking Neural Networks through Spatiotemporal Separation

Authors: Yiting Dong, Zhaofei Yu, Jianhao Ding, Zijie Xu, Tiejun Huang
Published: 2025-12-05
Categories: cs.NE
arXiv: https://arxiv.org/abs/2512.05472v1

Spiking Neural Networks (SNNs) are considered naturally suited for temporal processing, with membrane potential propagation widely regarded as the core temporal modeling mechanism. However, existing research lack analysis of its actual contributions in complex temporal tasks. We design Non-Stateful (NS) models progressively removing membrane propagation to quantify its stage-wise role. Experiments reveal a counterintuitive phenomenon: moderate removal in shallow or deep layers improves performance, while excessive removal causes collapse. We attribute this to spatio-temporal resource competition where neurons encode both semantics and dynamics within limited range, with temporal state consuming capacity for spatial learning. Based on this, we propose Spatial-Temporal Separable Network (STSep), decoupling residual blocks into independent spatial and temporal branches. The spatial branch focuses on semantic extraction while the temporal branch captures motion through explicit temporal differences. Experiments on Something-Something V2, UCF101, and HMDB51 show STSep achieves superior performance, with retrieval task and attention analysis confirming focus on motion rather than static appearance. This work provides new perspectives on SNNs' temporal mechanisms and an effective solution for spatiotemporal modeling in video understanding.

---

### How Ensemble Learning Balances Accuracy and Overfitting: A Bias-Variance Perspective on Tabular Data

Authors: Zubair Ahmed Mohammad
Published: 2025-12-05
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.05469v1

Ensemble models often achieve higher accuracy than single learners, but their ability to maintain small generalization gaps is not always well understood. This study examines how ensembles balance accuracy and overfitting across four tabular classification tasks: Breast Cancer, Heart Disease, Pima Diabetes, and Credit Card Fraud. Using repeated stratified cross validation with statistical significance testing, we compare linear models, a single decision tree, and nine ensemble methods. The results show that ensembles can reach high accuracy without large gaps by reducing variance through averaging or controlled boosting. On nearly linear and clean data, linear models already generalize well and ensembles offer little additional benefit. On datasets with meaningful nonlinear structure, tree based ensembles increase test accuracy by 5 to 7 points while keeping gaps below 3 percent. On noisy or highly imbalanced datasets, ensembles remain competitive but require regularization to avoid fitting noise or majority class patterns. We also compute simple dataset complexity indicators, such as linearity score, Fisher ratio, and noise estimate, which explain when ensembles are likely to control variance effectively. Overall, the study provides a clear view of how and when ensembles maintain high accuracy while keeping overfitting low, offering practical guidance for model selection in real world tabular applications.

---

### University Building Recognition Dataset in Thailand for the mission-oriented IoT sensor system

Authors: Takara Taniguchi, Yudai Ueda, Atsuya Muramatsu, Kohki Hashimoto, Ryo Yagi, Hideya Ochiai, Chaodit Aswakul
Published: 2025-12-05
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2512.05468v1

Many industrial sectors have been using of machine learning at inference mode on edge devices. Future directions show that training on edge devices is promising due to improvements in semiconductor performance. Wireless Ad Hoc Federated Learning (WAFL) has been proposed as a promising approach for collaborative learning with device-to-device communication among edges. In particular, WAFL with Vision Transformer (WAFL-ViT) has been tested on image recognition tasks with the UTokyo Building Recognition Dataset (UTBR). Since WAFL-ViT is a mission-oriented sensor system, it is essential to construct specific datasets by each mission. In our work, we have developed the Chulalongkorn University Building Recognition Dataset (CUBR), which is specialized for Chulalongkorn University as a case study in Thailand. Additionally, our results also demonstrate that training on WAFL scenarios achieves better accuracy than self-training scenarios. Dataset is available in https://github.com/jo2lxq/wafl/.

---

### Dynamic Alignment for Collective Agency: Toward a Scalable Self-Improving Framework for Open-Ended LLM Alignment

Authors: Panatchakorn Anantaprayoon, Nataliia Babina, Jad Tarifi, Nima Asgharbeygi
Published: 2025-12-05
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2512.05464v1

Large Language Models (LLMs) are typically aligned with human values using preference data or predefined principles such as helpfulness, honesty, and harmlessness. However, as AI systems progress toward Artificial General Intelligence (AGI) and Artificial Superintelligence (ASI), such value systems may become insufficient. In addition, human feedback-based alignment remains resource-intensive and difficult to scale. While AI-feedback-based self-improving alignment methods have been explored as a scalable alternative, they have largely remained constrained to conventional alignment values. In this work, we explore both a more holistic alignment objective and a scalable, self-improving alignment approach. Aiming to transcend conventional alignment norms, we introduce Collective Agency (CA)-a unified and open-ended alignment value that encourages integrated agentic capabilities. We also propose Dynamic Alignment-an alignment framework that enables an LLM to iteratively align itself. Dynamic Alignment comprises two key components: (1) automated training dataset generation with LLMs, and (2) a self-rewarding mechanism, where the policy model evaluates its own output candidates and assigns rewards for GRPO-based learning. Experimental results demonstrate that our approach successfully aligns the model to CA while preserving general NLP capabilities.

---

### Model Gateway: Management Platform for Model-Driven Drug Discovery

Authors: Yan-Shiun Wu, Sai Mahit Vaddadi, Zachary A. Rollins, Nathan A. Morin
Published: 2025-12-05
Categories: cs.SE, cs.DC, cs.LG, q-bio.QM
arXiv: https://arxiv.org/abs/2512.05462v2

Pharmaceutical drug discovery demands machine learning (ML) infrastructure that goes beyond general-purpose Machine Learning Operations (MLOps): inference-time composition of multiple models for multi-parameter optimization (MPO), version management for physics-based models without serialized ML artifacts, enterprise compound library precomputation, and governance structured around scientific organizational units rather than generic access controls. No existing commercial or open-source platform simultaneously addresses this full set of requirements. This paper presents the Model Gateway, a cloud-based platform for managing machine learning and scientific computational models across drug discovery pipelines, providing centralized version control, pharma-structured governance, asynchronous execution, consensus model orchestration, automated retraining, and a unified application programming interface (API) service for heterogeneous clients including molecular design suites and Large Language Model (LLM) agents. In production at Eli Lilly, the platform governs more than 200 deployed models spanning small molecule, peptide, and antibody modalities and serves more than five downstream applications across all phases of the Design-Make-Test-Analyze cycle.

---

### Knowing Your Uncertainty -- On the application of LLM in social sciences

Authors: Bolun Zhang, Linzhuo Li, Yunqi Chen, Qinlin Zhao, Zihan Zhu, Xiaoyuan Yi, Xing Xie
Published: 2025-12-05
Categories: cs.CY, cs.AI, cs.HC
arXiv: https://arxiv.org/abs/2512.05461v2

Large language models (LLMs) are rapidly being integrated into computational social science research, yet their blackboxed training and designed stochastic elements in inference pose unique challenges for scientific inquiry. This article argues that applying LLMs to social scientific tasks requires explicit assessment of uncertainty -- an expectation long established in both quantitative methodology in the social sciences and machine learning. We introduce a unified framework for evaluating LLM uncertainty based on Hill numbers, a family of diversity measures. By transforming existing uncertainty quantification (UQ) metrics into Hill numbers, the framework provides a common and intuitive scale for interpreting variation in LLM outputs while accommodating different notions of semantic similarity and different sensitivities to output distributions. We show how it might help the application of LLMs in social sciences through four empirical applications.

---

### Do We Really Even Need Data? A Modern Look at Drawing Inference with Predicted Data

Authors: Stephen Salerno, Kentaro Hoffman, Awan Afiaz, Anna Neufeld, Tyler H. McCormick, Jeffrey T. Leek
Published: 2025-12-05
Categories: stat.ML, cs.LG
arXiv: https://arxiv.org/abs/2512.05456v1

As artificial intelligence and machine learning tools become more accessible, and scientists face new obstacles to data collection (e.g., rising costs, declining survey response rates), researchers increasingly use predictions from pre-trained algorithms as substitutes for missing or unobserved data. Though appealing for financial and logistical reasons, using standard tools for inference can misrepresent the association between independent variables and the outcome of interest when the true, unobserved outcome is replaced by a predicted value. In this paper, we characterize the statistical challenges inherent to drawing inference with predicted data (IPD) and show that high predictive accuracy does not guarantee valid downstream inference. We show that all such failures reduce to statistical notions of (i) bias, when predictions systematically shift the estimand or distort relationships among variables, and (ii) variance, when uncertainty from the prediction model and the intrinsic variability of the true data are ignored. We then review recent methods for conducting IPD and discuss how this framework is deeply rooted in classical statistical theory. We then comment on some open questions and interesting avenues for future work in this area, and end with some comments on how to use predicted data in scientific studies that is both transparent and statistically principled.

---

### Parajudica: An RDF-Based Reasoner and Metamodel for Multi-Framework Context-Dependent Data Compliance Assessments

Authors: Luc Moreau, Alfred Rossi, Sophie Stalla-Bourdillon
Published: 2025-12-05
Categories: cs.DB, cs.AI, cs.CY, cs.LO
arXiv: https://arxiv.org/abs/2512.05453v1

Motivated by the challenges of implementing policy-based data access control (PBAC) under multiple simultaneously applicable compliance frameworks, we present Parajudica, an open, modular, and extensible RDF/SPARQL-based rule system for evaluating context-dependent data compliance status. We demonstrate the utility of this resource and accompanying metamodel through application to existing legal frameworks and industry standards, offering insights for comparative framework analysis. Applications include compliance policy enforcement, compliance monitoring, data discovery, and risk assessment.

---

### The Seeds of Scheming: Weakness of Will in the Building Blocks of Agentic Systems

Authors: Robert Yang
Published: 2025-12-05
Categories: cs.AI
arXiv: https://arxiv.org/abs/2512.05449v1

Large language models display a peculiar form of inconsistency: they "know" the correct answer but fail to act on it. In human philosophy, this tension between global judgment and local impulse is called akrasia, or weakness of will. We propose akrasia as a foundational concept for analyzing inconsistency and goal drift in agentic AI systems. To operationalize it, we introduce a preliminary version of the Akrasia Benchmark, currently a structured set of prompting conditions (Baseline (B), Synonym (S), Temporal (T), and Temptation (X)) that measures when a model's local response contradicts its own prior commitments. The benchmark enables quantitative comparison of "self-control" across model families, decoding strategies, and temptation types. Beyond single-model evaluation, we outline how micro-level akrasia may compound into macro-level instability in multi-agent systems that may be interpreted as "scheming" or deliberate misalignment. By reframing inconsistency as weakness of will, this work connects agentic behavior to classical theories of agency and provides an empirical bridge between philosophy, psychology, and the emerging science of agentic AI.

---

### TED-4DGS: Temporally Activated and Embedding-based Deformation for 4DGS Compression

Authors: Cheng-Yuan Ho, He-Bi Yang, Jui-Chiu Chiang, Yu-Lun Liu, Wen-Hsiao Peng
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.05446v1

Building on the success of 3D Gaussian Splatting (3DGS) in static 3D scene representation, its extension to dynamic scenes, commonly referred to as 4DGS or dynamic 3DGS, has attracted increasing attention. However, designing more compact and efficient deformation schemes together with rate-distortion-optimized compression strategies for dynamic 3DGS representations remains an underexplored area. Prior methods either rely on space-time 4DGS with overspecified, short-lived Gaussian primitives or on canonical 3DGS with deformation that lacks explicit temporal control. To address this, we present TED-4DGS, a temporally activated and embedding-based deformation scheme for rate-distortion-optimized 4DGS compression that unifies the strengths of both families. TED-4DGS is built on a sparse anchor-based 3DGS representation. Each canonical anchor is assigned learnable temporal-activation parameters to specify its appearance and disappearance transitions over time, while a lightweight per-anchor temporal embedding queries a shared deformation bank to produce anchor-specific deformation. For rate-distortion compression, we incorporate an implicit neural representation (INR)-based hyperprior to model anchor attribute distributions, along with a channel-wise autoregressive model to capture intra-anchor correlations. With these novel elements, our scheme achieves state-of-the-art rate-distortion performance on several real-world datasets. To the best of our knowledge, this work represents one of the first attempts to pursue a rate-distortion-optimized compression framework for dynamic 3DGS representations.

---

### IdealTSF: Can Non-Ideal Data Contribute to Enhancing the Performance of Time Series Forecasting Models?

Authors: Hua Wang, Jinghao Lu, Fan Zhang
Published: 2025-12-05
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.05442v2

Deep learning has shown strong performance in time series forecasting tasks. However, issues such as missing values and anomalies in sequential data hinder its further development in prediction tasks. Previous research has primarily focused on extracting feature information from sequence data or addressing these suboptimal data as positive samples for knowledge transfer. A more effective approach would be to leverage these non-ideal negative samples to enhance event prediction. In response, this study highlights the advantages of non-ideal negative samples and proposes the IdealTSF framework, which integrates both ideal positive and negative samples for time series forecasting. IdealTSF consists of three progressive steps: pretraining, training, and optimization. It first pretrains the model by extracting knowledge from negative sample data, then transforms the sequence data into ideal positive samples during training. Additionally, a negative optimization mechanism with adversarial disturbances is applied. Extensive experiments demonstrate that negative sample data unlocks significant potential within the basic attention architecture for time series forecasting. Therefore, IdealTSF is particularly well-suited for applications with noisy samples or low-quality data.

---

### BEAVER: An Efficient Deterministic LLM Verifier

Authors: Tarun Suresh, Nalin Wadhwa, Debangshu Banerjee, Gagandeep Singh
Published: 2025-12-05
Categories: cs.AI, cs.FL
arXiv: https://arxiv.org/abs/2512.05439v2

As large language models (LLMs) transition from research prototypes to production systems, practitioners often need reliable methods to verify model outputs and characterize tail risk for safe deployment. While sampling-based estimates provide an ad-hoc intuition of model behavior, they offer no sound guarantees. We present BEAVER, the first practical framework for computing deterministic, sound probability bounds on LLM satisfaction of safety properties. Given a prompt & any safety property, BEAVER systematically explores the model output space using novel Token trie and Frontier data structures, maintaining provably sound bounds at every iteration. We formalize the verification problem, prove soundness of our approach, and evaluate BEAVER on 4 safety properties across 12 open-weight LLMs. BEAVER identifies 2-3x more risky instances compared to baselines while taking 1/10 of the compute budget, surfacing tail risks that loose bounds and ad-hoc evaluation misses.

---

### EXR: An Interactive Immersive EHR Visualization in Extended Reality

Authors: Benoit Marteau, Shaun Q. Y. Tan, Jieru Li, Andrew Hornback, Yishan Zhong, Shaunna Wang, Christian Lowson, Jason Woloff, Joshua M. Pahys, Steven W. Hwang, Coleman Hilton, May D. Wang
Published: 2025-12-05
Categories: cs.HC, cs.CV, cs.LG, cs.MM
arXiv: https://arxiv.org/abs/2512.05438v1

This paper presents the design and implementation of an Extended Reality (XR) platform for immersive, interactive visualization of Electronic Health Records (EHRs). The system extends beyond conventional 2D interfaces by visualizing both structured and unstructured patient data into a shared 3D environment, enabling intuitive exploration and real-time collaboration. The modular infrastructure integrates FHIR-based EHR data with volumetric medical imaging and AI-generated segmentation, ensuring interoperability with modern healthcare systems. The platform's capabilities are demonstrated using synthetic EHR datasets and computed tomography (CT)-derived spine models processed through an AI-powered segmentation pipeline. This work suggests that such integrated XR solutions could form the foundation for next-generation clinical decision-support tools, where advanced data infrastructures are directly accessible in an interactive and spatially rich environment.

---

### Building Capacity for Artificial Intelligence in Africa: A Cross-Country Survey of Challenges and Governance Pathways

Authors: Jeffrey N. A. Aryee, Patrick Davies, Godfred A. Torsah, Mercy M. Apaw, Cyril D. Boateng, Sam M. Mwando, Chris Kwisanga, Eric Jobunga, Leonard K. Amekudzi
Published: 2025-12-05
Categories: cs.CY, cs.AI
arXiv: https://arxiv.org/abs/2512.05432v1

Artificial intelligence (AI) is transforming education and the workforce, but access to AI learning opportunities in Africa remains uneven. With rapid demographic shifts and growing labour market pressures, AI has become a strategic development priority, making the demand for relevant skills more urgent. This study investigates how universities and industries engage in shaping AI education and workforce preparation, drawing on survey responses from five African countries (Ghana, Namibia, Rwanda, Kenya and Zambia). The findings show broad recognition of AI importance but limited evidence of consistent engagement, practical training, or equitable access to resources. Most respondents who rated the AI component of their curriculum as very relevant reported being well prepared for jobs, but financial barriers, poor infrastructure, and weak communication limit participation, especially among students and underrepresented groups. Respondents highlighted internships, industry partnerships, and targeted support mechanisms as critical enablers, alongside the need for inclusive governance frameworks. The results showed both the growing awareness of AI's potential and the structural gaps that hinder its translation into workforce capacity. Strengthening university-industry collaboration and addressing barriers of access, funding, and policy are central to ensuring that AI contributes to equitable and sustainable development across the continent.

---

### ArtistMus: A Globally Diverse, Artist-Centric Benchmark for Retrieval-Augmented Music Question Answering

Authors: Daeyong Kwon, SeungHeon Doh, Juhan Nam
Published: 2025-12-05
Categories: cs.CL, cs.AI, cs.IR, cs.LG
arXiv: https://arxiv.org/abs/2512.05430v2

Recent advances in large language models (LLMs) have transformed open-domain question answering, yet their effectiveness in music-related reasoning remains limited due to sparse music knowledge in pretraining data. While music information retrieval and computational musicology have explored structured and multimodal understanding, few resources support factual and contextual music question answering (MQA) grounded in artist metadata or historical context. We introduce MusWikiDB, a vector database of 3.2M passages from 144K music-related Wikipedia pages, and ArtistMus, a benchmark of 1,000 questions on 500 diverse artists with metadata such as genre, debut year, and topic. These resources enable systematic evaluation of retrieval-augmented generation (RAG) for MQA. Experiments show that RAG markedly improves factual accuracy; open-source models gain up to +56.8 percentage points (for example, Qwen3 8B improves from 35.0 to 91.8), approaching proprietary model performance. RAG-style fine-tuning further boosts both factual recall and contextual reasoning, improving results on both in-domain and out-of-domain benchmarks. MusWikiDB also yields approximately 6 percentage points higher accuracy and 40% faster retrieval than a general-purpose Wikipedia corpus. We release MusWikiDB and ArtistMus to advance research in music information retrieval and domain-specific question answering, establishing a foundation for retrieval-augmented reasoning in culturally rich domains such as music.

---

### ParaUni: Enhance Generation in Unified Multimodal Model with Reinforcement-driven Hierarchical Parallel Information Interaction

Authors: Jiangtong Tan, Lin Liu, Jie Huanng, Xiaopeng Zhang, Qi Tian, Feng Zhao
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.05422v2

Unified multimodal models significantly improve visual generation by combining vision-language models (VLMs) with diffusion models. However, existing methods struggle to fully balance sufficient interaction and flexible implementation due to vast representation difference. Considering abundant and hierarchical information in VLM's layers from low-level details to high-level semantics, we propose \textbf{ParaUni}. It extracts features from variants VLM's layers in a \textbf{Para}llel way for comprehensive information interaction and retains a flexible separation architecture to enhance generation in \textbf{Uni}fied multimodal model. Concretely, visual features from all VLM's layers are fed in parallel into a Layer Integration Module (LIM), which efficiently integrates fine-grained details and semantic abstractions and provides the fused representation as a condition to the diffusion model. To further enhance performance, we reveal that these hierarchical layers respond unequally to different rewards in Reinforcement Learning (RL). Crucially, we design a Layer-wise Dynamic Adjustment Mechanism (LDAM) to facilitate multiple reward improvements that aligns the hierarchical properties of these layers using RL. Extensive experiments show ParaUni leverages complementary multi-layer features to substantially improve generation quality and shows strong potential for multiple reward advances during RL stages. Code is available at https://github.com/JosephTiTan/ParaUni.

---