# Research Papers — 2025-12 · bundle page 7 of 175

> Full-text markdown of papers #301–350 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/7
[← Previous bundle](https://aipod.dev/papers/archive/2025-12/page/6.md)
[Next bundle →](https://aipod.dev/papers/archive/2025-12/page/8.md)

### Training-Free Color-Aware Adversarial Diffusion Sanitization for Diffusion Stegomalware Defense at Security Gateways

Authors: Vladimir Frants, Sos Agaian
Published: 2025-12-30
Categories: cs.CR, cs.CV
arXiv: https://arxiv.org/abs/2512.24499v1

The rapid expansion of generative AI has normalized large-scale synthetic media creation, enabling new forms of covert communication. Recent generative steganography methods, particularly those based on diffusion models, can embed high-capacity payloads without fine-tuning or auxiliary decoders, creating significant challenges for detection and remediation. Coverless diffusion-based techniques are difficult to counter because they generate image carriers directly from secret data, enabling attackers to deliver stegomalware for command-and-control, payload staging, and data exfiltration while bypassing detectors that rely on cover-stego discrepancies. This work introduces Adversarial Diffusion Sanitization (ADS), a training-free defense for security gateways that neutralizes hidden payloads rather than detecting them. ADS employs an off-the-shelf pretrained denoiser as a differentiable proxy for diffusion-based decoders and incorporates a color-aware, quaternion-coupled update rule to reduce artifacts under strict distortion limits. Under a practical threat model and in evaluation against the state-of-the-art diffusion steganography method Pulsar, ADS drives decoder success rates to near zero with minimal perceptual impact. Results demonstrate that ADS provides a favorable security-utility trade-off compared to standard content transformations, offering an effective mitigation strategy against diffusion-driven steganography.

---

### What Drives Success in Physical Planning with Joint-Embedding Predictive World Models?

Authors: Basile Terver, Tsung-Yen Yang, Jean Ponce, Adrien Bardes, Yann LeCun
Published: 2025-12-30
Categories: cs.AI, cs.LG, cs.RO, stat.ML
arXiv: https://arxiv.org/abs/2512.24497v4

A long-standing challenge in AI is to develop agents capable of solving a wide range of physical tasks and generalizing to new, unseen tasks and environments. A popular recent approach involves training a world model from state-action trajectories and subsequently use it with a planning algorithm to solve new tasks. Planning is commonly performed in the input space, but a recent family of methods has introduced planning algorithms that optimize in the learned representation space of the world model, with the promise that abstracting irrelevant details yields more efficient planning. In this work, we characterize models from this family as JEPA-WMs and investigate the technical choices that make algorithms from this class work. We propose a comprehensive study of several key components with the objective of finding the optimal approach within the family. We conducted experiments using both simulated environments and real-world robotic data, and studied how the model architecture, the training objective, and the planning algorithm affect planning success. We combine our findings to propose a model that outperforms two established baselines, DINO-WM and V-JEPA-2-AC, in both navigation and manipulation tasks. Code, data and checkpoints are available at https://github.com/facebookresearch/jepa-wms.

---

### Automated Classification of First-Trimester Fetal Heart Views Using Ultrasound-Specific Self-Supervised Learning

Authors: Youssef Megahed, Aylin Erman, Robin Ducharme, Mark C. Walker, Steven Hawken, Adrian D. C. Chan
Published: 2025-12-30
Categories: eess.IV, cs.AI, cs.CV
arXiv: https://arxiv.org/abs/2512.24492v1

Congenital heart disease remains the most common congenital anomaly and a leading cause of neonatal morbidity and mortality. Although first-trimester fetal echocardiography offers an opportunity for earlier detection, automated analysis at this stage is challenging due to small cardiac structures, low signal-to-noise ratio, and substantial inter-operator variability. In this work, we evaluate a self-supervised ultrasound foundation model, USF-MAE, for first-trimester fetal heart view classification. USF-MAE is pretrained using masked autoencoding modelling on more than 370,000 unlabelled ultrasound images spanning over 40 anatomical regions and is subsequently fine-tuned for downstream classification. As a proof of concept, the pretrained Vision Transformer encoder was fine-tuned on an open-source dataset of 6,720 first-trimester fetal echocardiography images to classify five categories: aorta, atrioventricular flows, V sign, X sign, and Other. Model performance was benchmarked against supervised convolutional neural network baselines (ResNet-18 and ResNet-50) and a Vision Transformer (ViT-B/16) model pretrained on natural images (ImageNet-1k). All models were trained and evaluated using identical preprocessing, data splits, and optimization protocols. On an independent test set, USF-MAE achieved the highest performance across all evaluation metrics, with 90.57% accuracy, 91.15% precision, 90.57% recall, and 90.71% F1-score. This represents an improvement of +2.03% in accuracy and +1.98% in F1-score compared with the strongest baseline, ResNet-18. The proposed approach demonstrated robust performance without reliance on aggressive image preprocessing or region-of-interest cropping and showed improved discrimination of non-diagnostic frames.

---

### HOLOGRAPH: Active Causal Discovery via Sheaf-Theoretic Alignment of Large Language Model Priors

Authors: Hyunjun Kim
Published: 2025-12-30
Categories: cs.LG, cs.AI, stat.ME
arXiv: https://arxiv.org/abs/2512.24478v2

Causal discovery from observational data remains fundamentally limited by identifiability constraints. Recent work has explored leveraging Large Language Models (LLMs) as sources of prior causal knowledge, but existing approaches rely on heuristic integration that lacks theoretical grounding. We introduce HOLOGRAPH, a framework that formalizes LLM-guided causal discovery through sheaf theory--representing local causal beliefs as sections of a presheaf over variable subsets. Our key insight is that coherent global causal structure corresponds to the existence of a global section, while topological obstructions manifest as non-vanishing sheaf cohomology. We propose the Algebraic Latent Projection to handle hidden confounders and Natural Gradient Descent on the belief manifold for principled optimization. Experiments on synthetic and real-world benchmarks demonstrate that HOLOGRAPH provides rigorous mathematical foundations while achieving competitive performance on causal discovery tasks with 50-100 variables. Our sheaf-theoretic analysis reveals that while Identity, Transitivity, and Gluing axioms are satisfied to numerical precision (<10^{-6}), the Locality axiom fails for larger graphs, suggesting fundamental non-local coupling in latent variable projections. Code is available at (https://github.com/hyunjun1121/holograph)(https://github.com/hyunjun1121/holograph).

---

### F2IDiff: Real-world Image Super-resolution using Feature to Image Diffusion Foundation Model

Authors: Devendra K. Jangid, Ripon K. Saha, Dilshan Godaliyadda, Jing Li, Seok-Jun Lee, Hamid R. Sheikh
Published: 2025-12-30
Categories: cs.CV, cs.AI, eess.IV
arXiv: https://arxiv.org/abs/2512.24473v1

With the advent of Generative AI, Single Image Super-Resolution (SISR) quality has seen substantial improvement, as the strong priors learned by Text-2-Image Diffusion (T2IDiff) Foundation Models (FM) can bridge the gap between High-Resolution (HR) and Low-Resolution (LR) images. However, flagship smartphone cameras have been slow to adopt generative models because strong generation can lead to undesirable hallucinations. For substantially degraded LR images, as seen in academia, strong generation is required and hallucinations are more tolerable because of the wide gap between LR and HR images. In contrast, in consumer photography, the LR image has substantially higher fidelity, requiring only minimal hallucination-free generation. We hypothesize that generation in SISR is controlled by the stringency and richness of the FM's conditioning feature. First, text features are high level features, which often cannot describe subtle textures in an image. Additionally, Smartphone LR images are at least $12MP$, whereas SISR networks built on T2IDiff FM are designed to perform inference on much smaller images ($<1MP$). As a result, SISR inference has to be performed on small patches, which often cannot be accurately described by text feature. To address these shortcomings, we introduce an SISR network built on a FM with lower-level feature conditioning, specifically DINOv2 features, which we call a Feature-to-Image Diffusion (F2IDiff) Foundation Model (FM). Lower level features provide stricter conditioning while being rich descriptors of even small patches.

---

### Foundation models on the bridge: Semantic hazard detection and safety maneuvers for maritime autonomy with vision-language models

Authors: Kim Alexander Christensen, Andreas Gudahl Tufte, Alexey Gusev, Rohan Sinha, Milan Ganai, Ole Andreas Alsos, Marco Pavone, Martin Steinert
Published: 2025-12-30
Categories: cs.RO, cs.AI
arXiv: https://arxiv.org/abs/2512.24470v2

The draft IMO MASS Code requires autonomous and remotely supervised maritime vessels to detect departures from their operational design domain, enter a predefined fallback that notifies the operator, permit immediate human override, and avoid changing the voyage plan without approval. Meeting these obligations in the alert-to-takeover gap calls for a short-horizon, human-overridable fallback maneuver. Classical maritime autonomy stacks struggle when the correct action depends on meaning (e.g., diver-down flag means people in the water, fire close by means hazard). We argue (i) that vision-language models (VLMs) provide semantic awareness for such out-of-distribution situations, and (ii) that a fast-slow anomaly pipeline with a short-horizon, human-overridable fallback maneuver makes this practical in the handover window. We introduce Semantic Lookout, a camera-only, candidate-constrained VLM fallback maneuver selector that selects one cautious action (or station-keeping) from water-valid, world-anchored trajectories under continuous human authority. On 40 harbor scenes we measure per-call scene understanding and latency, alignment with human consensus (model majority-of-three voting), short-horizon risk-relief on fire hazard scenes, and an on-water alert->fallback maneuver->operator handover. Sub-10 s models retain most of the awareness of slower state-of-the-art models. The fallback maneuver selector outperforms geometry-only baselines and increases standoff distance on fire scenes. A field run verifies end-to-end operation. These results support VLMs as semantic fallback maneuver selectors compatible with the draft IMO MASS Code, within practical latency budgets, and motivate future work on domain-adapted, hybrid autonomy that pairs foundation-model semantics with multi-sensor bird's-eye-view perception and short-horizon replanning. Website: kimachristensen.github.io/bridge_policy

---

### Spectral and Spatial Graph Learning for Multispectral Solar Image Compression

Authors: Prasiddha Siwakoti, Atefeh Khoshkhahtinat, Piyush M. Mehta, Barbara J. Thompson, Michael S. F. Kirk, Daniel da Silva
Published: 2025-12-30
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2512.24463v1

High-fidelity compression of multispectral solar imagery remains challenging for space missions, where limited bandwidth must be balanced against preserving fine spectral and spatial details. We present a learned image compression framework tailored to solar observations, leveraging two complementary modules: (1) the Inter-Spectral Windowed Graph Embedding (iSWGE), which explicitly models inter-band relationships by representing spectral channels as graph nodes with learned edge features; and (2) the Windowed Spatial Graph Attention and Convolutional Block Attention (WSGA-C), which combines sparse graph attention with convolutional attention to reduce spatial redundancy and emphasize fine-scale structures. Evaluations on the SDOML dataset across six extreme ultraviolet (EUV) channels show that our approach achieves a 20.15%reduction in Mean Spectral Information Divergence (MSID), up to 1.09% PSNR improvement, and a 1.62% log transformed MS-SSIM gain over strong learned baselines, delivering sharper and spectrally faithful reconstructions at comparable bits-per-pixel rates. The code is publicly available at https://github.com/agyat4/sgraph .

---

### Align While Search: Belief-Guided Exploratory Inference for World-Grounded Embodied Agents

Authors: Seohui Bae, Jeonghye Kim, Youngchul Sung, Woohyung Lim
Published: 2025-12-30
Categories: cs.AI
arXiv: https://arxiv.org/abs/2512.24461v1

In this paper, we propose a test-time adaptive agent that performs exploratory inference through posterior-guided belief refinement without relying on gradient-based updates or additional training for LLM agent operating under partial observability. Our agent maintains an external structured belief over the environment state, iteratively updates it via action-conditioned observations, and selects actions by maximizing predicted information gain over the belief space. We estimate information gain using a lightweight LLM-based surrogate and assess world alignment through a novel reward that quantifies the consistency between posterior belief and ground-truth environment configuration. Experiments show that our method outperforms inference-time scaling baselines such as prompt-augmented or retrieval-enhanced LLMs, in aligning with latent world states with significantly lower integration overhead.

---

### IELTS Writing Revision Platform with Automated Essay Scoring and Adaptive Feedback

Authors: Titas Ramancauskas, Kotryna Ramancauske
Published: 2025-12-30
Categories: cs.CL, cs.HC
arXiv: https://arxiv.org/abs/2512.24460v1

This paper presents the design, development, and evaluation of a proposed revision platform assisting candidates for the International English Language Testing System (IELTS) writing exam. Traditional IELTS preparation methods lack personalised feedback, catered to the IELTS writing rubric. To address these shortcomings, the platform features an attractive user interface (UI), an Automated Essay Scoring system (AES), and targeted feedback tailored to candidates and the IELTS writing rubric. The platform architecture separates conversational guidance from a dedicated writing interface to reduce cognitive load and simulate exam conditions. Through iterative, Design-Based Research (DBR) cycles, the study progressed from rule-based to transformer-based with a regression head scoring, mounted with adaptive feedback. Early cycles (2-3) revealed fundamental limitations of rule-based approaches: mid-band compression, low accuracy, and negative $R^2$ values. DBR Cycle 4 implemented a DistilBERT transformer model with a regression head, yielding substantial improvements with MAE of 0.66 and positive $R^2$. This enabled Cycle 5's adaptive feedback implementation, which demonstrated statistically significant score improvements (mean +0.060 bands, p = 0.011, Cohen's d = 0.504), though effectiveness varied by revision strategy. Findings suggest automated feedback functions are most suited as a supplement to human instruction, with conservative surface-level corrections proving more reliable than aggressive structural interventions for IELTS preparation contexts. Challenges remain in assessing higher-band essays, and future work should incorporate longitudinal studies with real IELTS candidates and validation from official examiners.

---

### Cleaning English Abstracts of Scientific Publications

Authors: Michael E. Rose, Nils A. Herrmann, Sebastian Erhardt
Published: 2025-12-30
Categories: cs.CL
arXiv: https://arxiv.org/abs/2512.24459v1

Scientific abstracts are often used as proxies for the content and thematic focus of research publications. However, a significant share of published abstracts contains extraneous information-such as publisher copyright statements, section headings, author notes, registrations, and bibliometric or bibliographic metadata-that can distort downstream analyses, particularly those involving document similarity or textual embeddings. We introduce an open-source, easy-to-integrate language model designed to clean English-language scientific abstracts by automatically identifying and removing such clutter. We demonstrate that our model is both conservative and precise, alters similarity rankings of cleaned abstracts and improves information content of standard-length embeddings.

---

### Privacy-Preserving Semantic Communications via Multi-Task Learning and Adversarial Perturbations

Authors: Yalin E. Sagduyu, Tugba Erpek, Aylin Yener, Sennur Ulukus
Published: 2025-12-30
Categories: cs.NI, cs.AI, cs.CR, cs.IT, cs.LG
arXiv: https://arxiv.org/abs/2512.24452v1

Semantic communications conveys task-relevant meaning rather than focusing solely on message reconstruction, improving bandwidth efficiency and robustness for next-generation wireless systems. However, learned semantic representations can still leak sensitive information to unintended receivers (eavesdroppers). This paper presents a deep learning-based semantic communication framework that jointly supports multiple receiver tasks while explicitly limiting semantic leakage to an eavesdropper. The legitimate link employs a learned encoder at the transmitter, while the receiver trains decoders for semantic inference and data reconstruction. The security problem is formulated via an iterative min-max optimization in which an eavesdropper is trained to improve its semantic inference, while the legitimate transmitter-receiver pair is trained to preserve task performance while reducing the eavesdropper's success. We also introduce an auxiliary layer that superimposes a cooperative, adversarially crafted perturbation on the transmitted waveform to degrade semantic leakage to an eavesdropper. Performance is evaluated over Rayleigh fading channels with additive white Gaussian noise using MNIST and CIFAR-10 datasets. Semantic accuracy and reconstruction quality improve with increasing latent dimension, while the min-max mechanism reduces the eavesdropper's inference performance significantly without degrading the legitimate receiver. The perturbation layer is successful in reducing semantic leakage even when the legitimate link is trained only for its own task. This comprehensive framework motivates semantic communication designs with tunable, end-to-end privacy against adaptive adversaries in realistic wireless settings.

---

### PackKV: Reducing KV Cache Memory Footprint through LLM-Aware Lossy Compression

Authors: Bo Jiang, Taolue Yang, Youyuan Liu, Xubin He, Sheng Di, Sian Jin
Published: 2025-12-30
Categories: cs.DC, cs.AI
arXiv: https://arxiv.org/abs/2512.24449v2

Transformer-based large language models (LLMs) have demonstrated remarkable potential across a wide range of practical applications. However, long-context inference remains a significant challenge due to the substantial memory requirements of the key-value (KV) cache, which can scale to several gigabytes as sequence length and batch size increase. In this paper, we present \textbf{PackKV}, a generic and efficient KV cache management framework optimized for long-context generation. %, which synergistically supports both latency-critical and throughput-critical inference scenarios. PackKV introduces novel lossy compression techniques specifically tailored to the characteristics of KV cache data, featuring a careful co-design of compression algorithms and system architecture. Our approach is compatible with the dynamically growing nature of the KV cache while preserving high computational efficiency. Experimental results show that, under the same and minimum accuracy drop as state-of-the-art quantization methods, PackKV achieves, on average, \textbf{153.2}\% higher memory reduction rate for the K cache and \textbf{179.6}\% for the V cache. Furthermore, PackKV delivers extremely high execution throughput, effectively eliminating decompression overhead and accelerating the matrix-vector multiplication operation. Specifically, PackKV achieves an average throughput improvement of \textbf{75.7}\% for K and \textbf{171.7}\% for V across A100 and RTX Pro 6000 GPUs, compared to cuBLAS matrix-vector multiplication kernels, while demanding less GPU memory bandwidth. Code available on https://github.com/BoJiang03/PackKV

---

### Generative forecasting with joint probability models

Authors: Patrick Wyrod, Ashesh Chattopadhyay, Daniele Venturi
Published: 2025-12-30
Categories: cs.LG, physics.comp-ph
arXiv: https://arxiv.org/abs/2512.24446v1

Chaotic dynamical systems exhibit strong sensitivity to initial conditions and often contain unresolved multiscale processes, making deterministic forecasting fundamentally limited. Generative models offer an appealing alternative by learning distributions over plausible system evolutions; yet, most existing approaches focus on next-step conditional prediction rather than the structure of the underlying dynamics. In this work, we reframe forecasting as a fully generative problem by learning the joint probability distribution of lagged system states over short temporal windows and obtaining forecasts through marginalization. This new perspective allows the model to capture nonlinear temporal dependencies, represent multistep trajectory segments, and produce next-step predictions consistent with the learned joint distribution. We also introduce a general, model-agnostic training and inference framework for joint generative forecasting and show how it enables assessment of forecast robustness and reliability using three complementary uncertainty quantification metrics (ensemble variance, short-horizon autocorrelation, and cumulative Wasserstein drift), without access to ground truth. We evaluate the performance of the proposed method on two canonical chaotic dynamical systems, the Lorenz-63 system and the Kuramoto-Sivashinsky equation, and show that joint generative models yield improved short-term predictive skill, preserve attractor geometry, and achieve substantially more accurate long-range statistical behaviour than conventional conditional next-step models.

---

### Adaptive Learning Guided by Bias-Noise-Alignment Diagnostics

Authors: Akash Samanta, Sheldon Williamson
Published: 2025-12-30
Categories: cs.LG, eess.SY
arXiv: https://arxiv.org/abs/2512.24445v2

Learning systems deployed in nonstationary and safety-critical environments often suffer from instability, slow convergence, or brittle adaptation when learning dynamics evolve over time. While modern optimization, reinforcement learning, and meta-learning methods adapt to gradient statistics, they largely ignore the temporal structure of the error signal itself. This paper proposes a diagnostic-driven adaptive learning framework that explicitly models error evolution through a principled decomposition into bias, capturing persistent drift; noise, capturing stochastic variability; and alignment, capturing repeated directional excitation leading to overshoot. These diagnostics are computed online from lightweight statistics of loss or temporal-difference (TD) error trajectories and are independent of model architecture or task domain. We show that the proposed bias-noise-alignment decomposition provides a unifying control backbone for supervised optimization, actor-critic reinforcement learning, and learned optimizers. Within this framework, we introduce three diagnostic-driven instantiations: the Human-inspired Supervised Adaptive Optimizer (HSAO), Hybrid Error-Diagnostic Reinforcement Learning (HED-RL) for actor-critic methods, and the Meta-Learned Learning Policy (MLLP). Under standard smoothness assumptions, we establish bounded effective updates and stability properties for all cases. Representative diagnostic illustrations in actor-critic learning highlight how the proposed signals modulate adaptation in response to TD error structure. Overall, this work elevates error evolution to a first-class object in adaptive learning and provides an interpretable, lightweight foundation for reliable learning in dynamic environments.

---

### Sparse classification with positive-confidence data in high dimensions

Authors: The Tien Mai, Mai Anh Nguyen, Trung Nghia Nguyen
Published: 2025-12-30
Categories: cs.LG, stat.ML
arXiv: https://arxiv.org/abs/2512.24443v1

High-dimensional learning problems, where the number of features exceeds the sample size, often require sparse regularization for effective prediction and variable selection. While established for fully supervised data, these techniques remain underexplored in weak-supervision settings such as Positive-Confidence (Pconf) classification. Pconf learning utilizes only positive samples equipped with confidence scores, thereby avoiding the need for negative data. However, existing Pconf methods are ill-suited for high-dimensional regimes. This paper proposes a novel sparse-penalization framework for high-dimensional Pconf classification. We introduce estimators using convex (Lasso) and non-convex (SCAD, MCP) penalties to address shrinkage bias and improve feature recovery. Theoretically, we establish estimation and prediction error bounds for the L1-regularized Pconf estimator, proving it achieves near minimax-optimal sparse recovery rates under Restricted Strong Convexity condition. To solve the resulting composite objective, we develop an efficient proximal gradient algorithm. Extensive simulations demonstrate that our proposed methods achieve predictive performance and variable selection accuracy comparable to fully supervised approaches, effectively bridging the gap between weak supervision and high-dimensional statistics.

---

### Towards mechanistic understanding in a data-driven weather model: internal activations reveal interpretable physical features

Authors: Theodore MacMillan, Nicholas T. Ouellette
Published: 2025-12-30
Categories: physics.ao-ph, cs.LG, physics.comp-ph
arXiv: https://arxiv.org/abs/2512.24440v1

Large data-driven physics models like DeepMind's weather model GraphCast have empirically succeeded in parameterizing time operators for complex dynamical systems with an accuracy reaching or in some cases exceeding that of traditional physics-based solvers. Unfortunately, how these data-driven models perform computations is largely unknown and whether their internal representations are interpretable or physically consistent is an open question. Here, we adapt tools from interpretability research in Large Language Models to analyze intermediate computational layers in GraphCast, leveraging sparse autoencoders to discover interpretable features in the neuron space of the model. We uncover distinct features on a wide range of length and time scales that correspond to tropical cyclones, atmospheric rivers, diurnal and seasonal behavior, large-scale precipitation patterns, specific geographical coding, and sea-ice extent, among others. We further demonstrate how the precise abstraction of these features can be probed via interventions on the prediction steps of the model. As a case study, we sparsely modify a feature corresponding to tropical cyclones in GraphCast and observe interpretable and physically consistent modifications to evolving hurricanes. Such methods offer a window into the black-box behavior of data-driven physics models and are a step towards realizing their potential as trustworthy predictors and scientifically valuable tools for discovery.

---

### Exploring Compositionality in Vision Transformers using Wavelet Representations

Authors: Akshad Shyam Purushottamdas, Pranav K Nayak, Divya Mehul Rajparia, Deekshith Patel, Yashmitha Gogineni, Konda Reddy Mopuri, Sumohana S. Channappayya
Published: 2025-12-30
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.24438v1

While insights into the workings of the transformer model have largely emerged by analysing their behaviour on language tasks, this work investigates the representations learnt by the Vision Transformer (ViT) encoder through the lens of compositionality. We introduce a framework, analogous to prior work on measuring compositionality in representation learning, to test for compositionality in the ViT encoder. Crucial to drawing this analogy is the Discrete Wavelet Transform (DWT), which is a simple yet effective tool for obtaining input-dependent primitives in the vision setting. By examining the ability of composed representations to reproduce original image representations, we empirically test the extent to which compositionality is respected in the representation space. Our findings show that primitives from a one-level DWT decomposition produce encoder representations that approximately compose in latent space, offering a new perspective on how ViTs structure information.

---

### Virasoro Symmetry in Neural Network Field Theories

Authors: Brandon Robinson
Published: 2025-12-30
Categories: hep-th, cs.LG
arXiv: https://arxiv.org/abs/2512.24420v3

Neural Network Field Theories (NN-FTs) typically describe Generalized Free Fields that lack a local stress-energy tensor in two dimensions, obstructing the realization of Virasoro symmetry. We present the ``Log-Kernel'' (LK) architecture, which enforces local conformal symmetry via a specific rotation-invariant spectral prior $p(k) \propto |k|^{-2}$. We analytically derive the emergence of the Virasoro algebra from the statistics of the neural ensemble. We validate this construction through numerical simulation, computing the central charge $c_{exp} = 0.9958 \pm 0.0196$ (theoretical $c=1$) and confirming the scaling dimensions of vertex operators. Furthermore, we demonstrate that finite-width corrections generate interactions scaling as $1/N$. Finally, we extend the framework to include fermions and boundary conditions, realizing the super-Virasoro algebra. We verify the $\mathcal{N}=1$ super-Virasoro algebra by measuring the supercurrent correlator to $96\%$ accuracy. We further demonstrate conformal boundary conditions on the upper half-plane, achieving 99\% agreement for boundary fermion and boson propagators.

---

### AI-Driven Evaluation of Surgical Skill via Action Recognition

Authors: Yan Meng, Daniel A. Donoho, Marcelle Altshuler, Omar Arnaout
Published: 2025-12-30
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.24411v1

The development of effective training and evaluation strategies is critical. Conventional methods for assessing surgical proficiency typically rely on expert supervision, either through onsite observation or retrospective analysis of recorded procedures. However, these approaches are inherently subjective, susceptible to inter-rater variability, and require substantial time and effort from expert surgeons. These demands are often impractical in low- and middle-income countries, thereby limiting the scalability and consistency of such methods across training programs. To address these limitations, we propose a novel AI-driven framework for the automated assessment of microanastomosis performance. The system integrates a video transformer architecture based on TimeSformer, improved with hierarchical temporal attention and weighted spatial attention mechanisms, to achieve accurate action recognition within surgical videos. Fine-grained motion features are then extracted using a YOLO-based object detection and tracking method, allowing for detailed analysis of instrument kinematics. Performance is evaluated along five aspects of microanastomosis skill, including overall action execution, motion quality during procedure-critical actions, and general instrument handling. Experimental validation using a dataset of 58 expert-annotated videos demonstrates the effectiveness of the system, achieving 87.7% frame-level accuracy in action segmentation that increased to 93.62% with post-processing, and an average classification accuracy of 76% in replicating expert assessments across all skill aspects. These findings highlight the system's potential to provide objective, consistent, and interpretable feedback, thereby enabling more standardized, data-driven training and evaluation in surgical education.

---

### Comparing Approaches to Automatic Summarization in Less-Resourced Languages

Authors: Chester Palen-Michel, Constantine Lignos
Published: 2025-12-30
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2512.24410v1

Automatic text summarization has achieved high performance in high-resourced languages like English, but comparatively less attention has been given to summarization in less-resourced languages. This work compares a variety of different approaches to summarization from zero-shot prompting of LLMs large and small to fine-tuning smaller models like mT5 with and without three data augmentation approaches and multilingual transfer. We also explore an LLM translation pipeline approach, translating from the source language to English, summarizing and translating back. Evaluating with five different metrics, we find that there is variation across LLMs in their performance across similar parameter sizes, that our multilingual fine-tuned mT5 baseline outperforms most other approaches including zero-shot LLM performance for most metrics, and that LLM as judge may be less reliable on less-resourced languages.

---

### DyStream: Streaming Dyadic Talking Heads Generation via Flow Matching-based Autoregressive Model

Authors: Bohong Chen, Haiyang Liu
Published: 2025-12-30
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.24408v2

Generating realistic, dyadic talking head video requires ultra-low latency. Existing chunk-based methods require full non-causal context windows, introducing significant delays. This high latency critically prevents the immediate, non-verbal feedback required for a realistic listener. To address this, we present DyStream, a flow matching-based autoregressive model that could generate video in real-time from both speaker and listener audio. Our method contains two key designs: (1) we adopt a stream-friendly autoregressive framework with flow-matching heads for probabilistic modeling, and (2) We propose a causal encoder enhanced by a lookahead module to incorporate short future context (e.g., 60 ms) to improve quality while maintaining low latency. Our analysis shows this simple-and-effective method significantly surpass alternative causal strategies, including distillation and generative encoder. Extensive experiments show that DyStream could generate video within 34 ms per frame, guaranteeing the entire system latency remains under 100 ms. Besides, it achieves state-of-the-art lip-sync quality, with offline and online LipSync Confidence scores of 8.13 and 7.61 on HDTF, respectively. The model, weights and codes are available.

---

### Efficient Inference for Inverse Reinforcement Learning and Dynamic Discrete Choice Models

Authors: Lars van der Laan, Aurélien Bibaut, Nathan Kallus
Published: 2025-12-30
Categories: cs.LG, math.ST
arXiv: https://arxiv.org/abs/2512.24407v2

In many sequential decision-making problems, researchers observe actions but not the rewards that drive behavior, yet still wish to evaluate and compare counterfactual policies. Inverse reinforcement learning (IRL) and dynamic discrete choice (DDC) models address this setting by positing an optimality model that links latent rewards to observed actions. Existing flexible IRL methods allow rich reward representations but typically do not provide valid inference, whereas classical DDC methods support inference only under restrictive parametric structure. We develop a semiparametric framework for debiased inverse reinforcement learning in maximum-entropy IRL and Gumbel-shock DDC models. Our key identification result is that the log-behavior policy can be treated as a pseudo-reward: it point-identifies policy value differences and, under a normalization constraint, the reward itself. This reduces inference on reward-dependent estimands to inference on smooth functionals of the behavior policy and transition kernel. We establish pathwise differentiability, derive efficient influence functions, and construct automatic debiased machine-learning estimators that permit flexible nuisance estimation while attaining $\sqrt{n}$-consistency, asymptotic normality, and semiparametric efficiency. The result is a computationally tractable framework for valid uncertainty quantification in flexible IRL and DDC models.

---

### Lifting Vision: Ground to Aerial Localization with Reasoning Guided Planning

Authors: Soham Pahari, M. Srinivas
Published: 2025-12-30
Categories: cs.LG, cs.CV
arXiv: https://arxiv.org/abs/2512.24404v1

Multimodal intelligence development recently show strong progress in visual understanding and high level reasoning. Though, most reasoning system still reply on textual information as the main medium for inference. This limit their effectiveness in spatial tasks such as visual navigation and geo-localization. This work discuss about the potential scope of this field and eventually propose an idea visual reasoning paradigm Geo-Consistent Visual Planning, our introduced framework called Visual Reasoning for Localization, or ViReLoc, which performs planning and localization using only visual representations. The proposed framework learns spatial dependencies and geometric relations that text based reasoning often suffer to understand. By encoding step by step inference in the visual domain and optimizing with reinforcement based objectives, ViReLoc plans routes between two given ground images. The system also integrates contrastive learning and adaptive feature interaction to align cross view perspectives and reduce viewpoint differences. Experiments across diverse navigation and localization scenarios show consistent improvements in spatial reasoning accuracy and cross view retrieval performance. These results establish visual reasoning as a strong complementary approach for navigation and localization, and show that such tasks can be performed without real time global positioning system data, leading to more secure navigation solutions.

---

### Fast and Realistic Automated Scenario Simulations and Reporting for an Autonomous Racing Stack

Authors: Giovanni Lambertini, Matteo Pini, Eugenio Mascaro, Francesco Moretti, Ayoub Raji, Marko Bertogna
Published: 2025-12-30
Categories: cs.RO, cs.AI, cs.SE, eess.SY
arXiv: https://arxiv.org/abs/2512.24402v1

In this paper, we describe the automated simulation and reporting pipeline implemented for our autonomous racing stack, ur.autopilot. The backbone of the simulation is based on a high-fidelity model of the vehicle interfaced as a Functional Mockup Unit (FMU). The pipeline can execute the software stack and the simulation up to three times faster than real-time, locally or on GitHub for Continuous Integration/- Continuous Delivery (CI/CD). As the most important input of the pipeline, there is a set of running scenarios. Each scenario allows the initialization of the ego vehicle in different initial conditions (position and speed), as well as the initialization of any other configuration of the stack. This functionality is essential to validate efficiently critical modules, like the one responsible for high-speed overtaking maneuvers or localization, which are among the most challenging aspects of autonomous racing. Moreover, we describe how we implemented a fault injection module, capable of introducing sensor delays and perturbations as well as modifying outputs of any node of the stack. Finally, we describe the design of our automated reporting process, aimed at maximizing the effectiveness of the simulation analysis.

---

### FAST-IDS: A Fast Two-Stage Intrusion Detection System with Hybrid Compression for Real-Time Threat Detection in Connected and Autonomous Vehicles

Authors: Devika S, Vishnu Hari, Pratik Narang, Tejasvi Alladi, Vinay Chamola
Published: 2025-12-30
Categories: cs.CR, cs.AI
arXiv: https://arxiv.org/abs/2512.24391v1

We have implemented a multi-stage IDS for CAVs that can be deployed to resourec-constrained environments after hybrid model compression.

---

### RedunCut: Measurement-Driven Sampling and Accuracy Performance Modeling for Low-Cost Live Video Analytics

Authors: Gur-Eyal Sela, Kumar Krishna Agrawal, Bharathan Balaji, Joseph Gonzalez, Ion Stoica
Published: 2025-12-30
Categories: cs.CV, cs.DC
arXiv: https://arxiv.org/abs/2512.24386v1

Live video analytics (LVA) runs continuously across massive camera fleets, but inference cost with modern vision models remains high. To address this, dynamic model size selection (DMSS) is an attractive approach: it is content-aware but treats models as black boxes, and could potentially reduce cost by up to 10x without model retraining or modification. Without ground truth labels at runtime, we observe that DMSS methods use two stages per segment: (i) sampling a few models to calculate prediction statistics (e.g., confidences), then (ii) selection of the model size from those statistics. Prior systems fail to generalize to diverse workloads, particularly to mobile videos and lower accuracy targets. We identify that the failure modes stem from inefficient sampling whose cost exceeds its benefit, and inaccurate per-segment accuracy prediction. In this work, we present RedunCut, a new DMSS system that addresses both: It uses a measurement-driven planner that estimates the cost-benefit tradeoff of sampling, and a lightweight, data-driven performance model to improve accuracy prediction. Across road-vehicle, drone, and surveillance videos and multiple model families and tasks, RedunCut reduces compute cost by 14-62% at fixed accuracy and remains robust to limited historical data and to drift.

---

### Forging Spatial Intelligence: A Roadmap of Multi-Modal Data Pre-Training for Autonomous Systems

Authors: Song Wang, Lingdong Kong, Xiaolu Liu, Hao Shi, Wentong Li, Jianke Zhu, Steven C. H. Hoi
Published: 2025-12-30
Categories: cs.CV, cs.RO
arXiv: https://arxiv.org/abs/2512.24385v2

The rapid advancement of autonomous systems, including self-driving vehicles and drones, has intensified the need to forge true Spatial Intelligence from multi-modal onboard sensor data. While foundation models excel in single-modal contexts, integrating their capabilities across diverse sensors like cameras and LiDAR to create a unified understanding remains a formidable challenge. This paper presents a comprehensive framework for multi-modal pre-training, identifying the core set of techniques driving progress toward this goal. We dissect the interplay between foundational sensor characteristics and learning strategies, evaluating the role of platform-specific datasets in enabling these advancements. Our central contribution is the formulation of a unified taxonomy for pre-training paradigms: ranging from single-modality baselines to sophisticated unified frameworks that learn holistic representations for advanced tasks like 3D object detection and semantic occupancy prediction. Furthermore, we investigate the integration of textual inputs and occupancy representations to facilitate open-world perception and planning. Finally, we identify critical bottlenecks, such as computational efficiency and model scalability, and propose a roadmap toward general-purpose multi-modal foundation models capable of achieving robust Spatial Intelligence for real-world deployment.

---

### Geometric Multi-Session Map Merging with Learned Local Descriptors

Authors: Yanlong Ma, Nakul S. Joshi, Christa S. Robison, Philip R. Osteen, Brett T. Lopez
Published: 2025-12-30
Categories: cs.RO, cs.CV
arXiv: https://arxiv.org/abs/2512.24384v1

Multi-session map merging is crucial for extended autonomous operations in large-scale environments. In this paper, we present GMLD, a learning-based local descriptor framework for large-scale multi-session point cloud map merging that systematically aligns maps collected across different sessions with overlapping regions. The proposed framework employs a keypoint-aware encoder and a plane-based geometric transformer to extract discriminative features for loop closure detection and relative pose estimation. To further improve global consistency, we include inter-session scan matching cost factors in the factor-graph optimization stage. We evaluate our framework on the public datasets, as well as self-collected data from diverse environments. The results show accurate and robust map merging with low error, and the learned features deliver strong performance in both loop closure detection and relative pose estimation.

---

### Tubular Riemannian Laplace Approximations for Bayesian Neural Networks

Authors: Rodrigo Pereira David
Published: 2025-12-30
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.24381v2

Laplace approximations are among the simplest and most practical methods for approximate Bayesian inference in neural networks, yet their Euclidean formulation struggles with the highly anisotropic, curved loss surfaces and large symmetry groups that characterize modern deep models. Recent work has proposed Riemannian and geometric Gaussian approximations to adapt to this structure. Building on these ideas, we introduce the Tubular Riemannian Laplace (TRL) approximation. TRL explicitly models the posterior as a probabilistic tube that follows a low-loss valley induced by functional symmetries, using a Fisher/Gauss-Newton metric to separate prior-dominated tangential uncertainty from data-dominated transverse uncertainty. We interpret TRL as a scalable reparametrised Gaussian approximation that utilizes implicit curvature estimates to operate in high-dimensional parameter spaces. Our empirical evaluation on ResNet-18 (CIFAR-10 and CIFAR-100) demonstrates that TRL achieves excellent calibration, matching or exceeding the reliability of Deep Ensembles (in terms of ECE) while requiring only a fraction (1/5) of the training cost. TRL effectively bridges the gap between single-model efficiency and ensemble-grade reliability.

---

### Implicit score matching meets denoising score matching: improved rates of convergence and log-density Hessian estimation

Authors: Konstantin Yakovlev, Anna Markovich, Nikita Puchkin
Published: 2025-12-30
Categories: math.ST, cs.LG, stat.ML
arXiv: https://arxiv.org/abs/2512.24378v1

We study the problem of estimating the score function using both implicit score matching and denoising score matching. Assuming that the data distribution exhibiting a low-dimensional structure, we prove that implicit score matching is able not only to adapt to the intrinsic dimension, but also to achieve the same rates of convergence as denoising score matching in terms of the sample size. Furthermore, we demonstrate that both methods allow us to estimate log-density Hessians without the curse of dimensionality by simple differentiation. This justifies convergence of ODE-based samplers for generative diffusion models. Our approach is based on Gagliardo-Nirenberg-type inequalities relating weighted $L^2$-norms of smooth functions and their derivatives.

---

### Skim-Aware Contrastive Learning for Efficient Document Representation

Authors: Waheed Ahmed Abro, Zied Bouraoui
Published: 2025-12-30
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2512.24373v1

Although transformer-based models have shown strong performance in word- and sentence-level tasks, effectively representing long documents, especially in fields like law and medicine, remains difficult. Sparse attention mechanisms can handle longer inputs, but are resource-intensive and often fail to capture full-document context. Hierarchical transformer models offer better efficiency but do not clearly explain how they relate different sections of a document. In contrast, humans often skim texts, focusing on important sections to understand the overall message. Drawing from this human strategy, we introduce a new self-supervised contrastive learning framework that enhances long document representation. Our method randomly masks a section of the document and uses a natural language inference (NLI)-based contrastive objective to align it with relevant parts while distancing it from unrelated ones. This mimics how humans synthesize information, resulting in representations that are both richer and more computationally efficient. Experiments on legal and biomedical texts confirm significant gains in both accuracy and efficiency.

---

### A Critical Assessment of PINNs and Operator Learning for Geotechnical Engineering

Authors: Krishna Kumar
Published: 2025-12-30
Categories: physics.geo-ph, cs.LG
arXiv: https://arxiv.org/abs/2512.24365v2

Scientific machine learning (SciML) offers neural-network alternatives to numerical workflows in geotechnical engineering. This paper benchmarks multi-layer perceptrons (MLPs), physics-informed neural networks (PINNs), deep operator networks (DeepONet), and graph network simulators (GNS) against finite-difference and particle-based references on geotechnical benchmarks, and compares PINN inversion with automatic differentiation (AD) through a conventional solver. We evaluate each method for extrapolation, training, and inference cost, transfer across problem instances, and physics accuracy. An MLP trained on two years of Terzaghi consolidation fits the data, but at year ten predicts ~290 mm with ReLU and ~60 mm with tanh or sigmoid, against a reference of 99.3 mm. A PINN on a damped oscillator with a time domain inside (0,1) matches the closed form within that interval but fails outside, since the residual constrains the fit only where it is sampled. For the 1D wave equation, PINN training is ~96,000 times slower than finite-difference methods and less accurate. DeepONet avoids PINN retraining, yet for the beam on elastic foundation, its training cost equals ~1.8 million finite-difference solves, and inference is slower per query than the direct solver. GNS improves geometric transfer through local particle interactions, though formulations still need trajectories, large training sets, and substantial memory. In the inverse wave benchmark, AD through the finite-difference solver recovers the material profile in seconds with ~1% error. The results support a cautious role for SciML. Neural networks suit interpolation and pattern recognition inside validated domains, while inverse analysis should first try differentiable physics-based solvers when a reliable forward solver exists.

---

### FedSecureFormer: A Fast, Federated and Secure Transformer Framework for Lightweight Intrusion Detection in Connected and Autonomous Vehicles

Authors: Devika S, Vishnu Hari, Pratik Narang, Tejasvi Alladi, F. Richard Yu
Published: 2025-12-30
Categories: cs.CR, cs.AI
arXiv: https://arxiv.org/abs/2512.24345v1

This works presents an encoder-only transformer built with minimum layers for intrusion detection in the domain of Connected and Autonomous Vehicles using Federated Learning.

---

### DermaVQA-DAS: Dermatology Assessment Schema (DAS) & Datasets for Closed-Ended Question Answering & Segmentation in Patient-Generated Dermatology Images

Authors: Wen-wai Yim, Yujuan Fu, Asma Ben Abacha, Meliha Yetisgen, Noel Codella, Roberto Andres Novoa, Josep Malvehy
Published: 2025-12-30
Categories: cs.CV, cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2512.24340v1

Recent advances in dermatological image analysis have been driven by large-scale annotated datasets; however, most existing benchmarks focus on dermatoscopic images and lack patient-authored queries and clinical context, limiting their applicability to patient-centered care. To address this gap, we introduce DermaVQA-DAS, an extension of the DermaVQA dataset that supports two complementary tasks: closed-ended question answering (QA) and dermatological lesion segmentation. Central to this work is the Dermatology Assessment Schema (DAS), a novel expert-developed framework that systematically captures clinically meaningful dermatological features in a structured and standardized form. DAS comprises 36 high-level and 27 fine-grained assessment questions, with multiple-choice options in English and Chinese. Leveraging DAS, we provide expert-annotated datasets for both closed QA and segmentation and benchmark state-of-the-art multimodal models. For segmentation, we evaluate multiple prompting strategies and show that prompt design impacts performance: the default prompt achieves the best results under Mean-of-Max and Mean-of-Mean evaluation aggregation schemes, while an augmented prompt incorporating both patient query title and content yields the highest performance under majority-vote-based microscore evaluation, achieving a Jaccard index of 0.395 and a Dice score of 0.566 with BiomedParse. For closed-ended QA, overall performance is strong across models, with average accuracies ranging from 0.729 to 0.798; o3 achieves the best overall accuracy (0.798), closely followed by GPT-4.1 (0.796), while Gemini-1.5-Pro shows competitive performance within the Gemini family (0.783). We publicly release DermaVQA-DAS, the DAS schema, and evaluation protocols to support and accelerate future research in patient-centered dermatological vision-language modeling (https://osf.io/72rp3).

---

### The Mechanics of CNN Filtering with Rectification

Authors: Liam Frija-Altarac, Matthew Toews
Published: 2025-12-30
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.24338v2

This paper proposes elementary information mechanics as a new model for understanding the mechanical properties of convolutional filtering with rectification, inspired by physical theories of special relativity and quantum mechanics. We consider kernels decomposed into orthogonal even and odd components. Even components cause image content to diffuse isotropically while preserving the center of mass, analogously to rest or potential energy with zero net momentum. Odd kernels cause directional displacement of the center of mass, analogously to kinetic energy with non-zero momentum. The speed of information displacement is linearly related to the ratio of odd vs total kernel energy. Even-Odd properties are analyzed in the spectral domain via the discrete cosine transform (DCT), where the structure of small convolutional filters (e.g. $3 \times 3$ pixels) is dominated by low-frequency bases, specifically the DC $Σ$ and gradient components $\nabla$, which define the fundamental modes of information propagation. To our knowledge, this is the first work demonstrating the link between information processing in generic CNNs and the energy-momentum relation, a cornerstone of modern relativistic physics.

---

### OptiVote: Non-Coherent FSO Over-the-Air Majority Vote for Communication-Efficient Distributed Federated Learning in Space Data Centers

Authors: Anbang Zhang, Chenyuan Feng, Wai Ho Mow, Jia Ye, Shuaishuai Guo, Geyong Min, Tony Q. S. Quek
Published: 2025-12-30
Categories: eess.SP, cs.LG
arXiv: https://arxiv.org/abs/2512.24334v2

The rapid deployment of mega-constellations is driving the long-term vision of space data centers (SDCs), where interconnected satellites form in-orbit distributed computing and learning infrastructures. Enabling distributed federated learning in such systems is challenging because iterative training requires frequent aggregation over inter-satellite links that are bandwidth- and energy-constrained, and the link conditions can be highly dynamic. In this work, we exploit over-the-air computation (AirComp) as an in-network aggregation primitive. However, conventional coherent AirComp relies on stringent phase alignment, which is difficult to maintain in space environments due to satellite jitter and Doppler effects. To overcome this limitation, we propose OptiVote, a robust and communication-efficient non-coherent free-space optical (FSO) AirComp framework for federated learning toward Space Data Centers. OptiVote integrates sign stochastic gradient descent (signSGD) with a majority-vote (MV) aggregation principle and pulse-position modulation (PPM), where each satellite conveys local gradient signs by activating orthogonal PPM time slots. The aggregation node performs MV detection via non-coherent energy accumulation, transforming phase-sensitive field superposition into phase-agnostic optical intensity combining, thereby eliminating the need for precise phase synchronization and improving resilience under dynamic impairments. To mitigate aggregation bias induced by heterogeneous FSO channels, we further develop an importance-aware, channel state information (CSI)-free dynamic power control scheme that balances received energies without additional signaling. We provide theoretical analysis by characterizing the aggregate error probability under statistical FSO channels and establishing convergence guarantees for non-convex objectives.

---

### Spatial-aware Vision Language Model for Autonomous Driving

Authors: Weijie Wei, Zhipeng Luo, Ling Feng, Venice Erin Liong
Published: 2025-12-30
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.24331v2

While Vision-Language Models (VLMs) show significant promise for end-to-end autonomous driving by leveraging the common sense embedded in language models, their reliance on 2D image cues for complex scene understanding and decision-making presents a critical bottleneck for safety and reliability. Current image-based methods struggle with accurate metric spatial reasoning and geometric inference, leading to unreliable driving policies. To bridge this gap, we propose LVLDrive (LiDAR-Vision-Language), a novel framework specifically designed to upgrade existing VLMs with robust 3D metric spatial understanding for autonomous driving by incoperating LiDAR point cloud as an extra input modality. A key challenge lies in mitigating the catastrophic disturbance introduced by disparate 3D data to the pre-trained VLMs. To this end, we introduce a Gradual Fusion Q-Former that incrementally injects LiDAR features, ensuring the stability and preservation of the VLM's existing knowledge base. Furthermore, we develop a spatial-aware question-answering (SA-QA) dataset to explicitly teach the model advanced 3D perception and reasoning capabilities. Extensive experiments on driving benchmarks demonstrate that LVLDrive achieves superior performance compared to vision-only counterparts across scene understanding, metric spatial perception, and reliable driving decision-making. Our work highlights the necessity of explicit 3D metric data for building trustworthy VLM-based autonomous systems.

---

### SenseNova-MARS: Empowering Multimodal Agentic Reasoning and Search via Reinforcement Learning

Authors: Yong Xien Chng, Tao Hu, Wenwen Tong, Xueheng Li, Jiandong Chen, Haojia Yu, Jiefan Lu, Hewei Guo, Hanming Deng, Chengjun Xie, Gao Huang, Dahua Lin et al.
Published: 2025-12-30
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.24330v2

While Vision-Language Models (VLMs) can solve complex tasks through agentic reasoning, their capabilities remain largely constrained to text-oriented chain-of-thought or isolated tool invocation. They fail to exhibit the human-like proficiency required to seamlessly interleave dynamic tool manipulation with continuous reasoning, particularly in knowledge-intensive and visually complex scenarios that demand coordinated external tools such as search and image cropping. In this work, we introduce SenseNova-MARS, a novel Multimodal Agentic Reasoning and Search framework that empowers VLMs with interleaved visual reasoning and tool-use capabilities via reinforcement learning (RL). Specifically, SenseNova-MARS dynamically integrates the image search, text search, and image crop tools to tackle fine-grained and knowledge-intensive visual understanding challenges. In the RL stage, we propose the Batch-Normalized Group Sequence Policy Optimization (BN-GSPO) algorithm to improve the training stability and advance the model's ability to invoke tools and reason effectively. To comprehensively evaluate the agentic VLMs on complex visual tasks, we introduce the HR-MMSearch benchmark, the first search-oriented benchmark composed of high-resolution images with knowledge-intensive and search-driven questions. Experiments demonstrate that SenseNova-MARS achieves state-of-the-art performance on open-source search and fine-grained image understanding benchmarks. Specifically, on search-oriented benchmarks, SenseNova-MARS-32B scores 74.3 on MMSearch and 54.4 on HR-MMSearch, surpassing proprietary models such as Gemini-3-Pro and GPT-5.2. SenseNova-MARS represents a promising step toward agentic VLMs by providing effective and robust tool-use capabilities. To facilitate further research in this field, we will release all code, models, and datasets.

---

### World model inspired sarcasm reasoning with large language model agents

Authors: Keito Inoshita, Shinnosuke Mizuno
Published: 2025-12-30
Categories: cs.CL
arXiv: https://arxiv.org/abs/2512.24329v1

Sarcasm understanding is a challenging problem in natural language processing, as it requires capturing the discrepancy between the surface meaning of an utterance and the speaker's intentions as well as the surrounding social context. Although recent advances in deep learning and Large Language Models (LLMs) have substantially improved performance, most existing approaches still rely on black-box predictions of a single model, making it difficult to structurally explain the cognitive factors underlying sarcasm. Moreover, while sarcasm often emerges as a mismatch between semantic evaluation and normative expectations or intentions, frameworks that explicitly decompose and model these components remain limited. In this work, we reformulate sarcasm understanding as a world model inspired reasoning process and propose World Model inspired SArcasm Reasoning (WM-SAR), which decomposes literal meaning, context, normative expectation, and intention into specialized LLM-based agents. The discrepancy between literal evaluation and normative expectation is explicitly quantified as a deterministic inconsistency score, and together with an intention score, these signals are integrated by a lightweight Logistic Regression model to infer the final sarcasm probability. This design leverages the reasoning capability of LLMs while maintaining an interpretable numerical decision structure. Experiments on representative sarcasm detection benchmarks show that WM-SAR consistently outperforms existing deep learning and LLM-based methods. Ablation studies and case analyses further demonstrate that integrating semantic inconsistency and intention reasoning is essential for effective sarcasm detection, achieving both strong performance and high interpretability.

---

### Topological Spatial Graph Coarsening

Authors: Anna Calissano, Etienne Lasalle
Published: 2025-12-30
Categories: stat.ML, cs.CG, cs.LG
arXiv: https://arxiv.org/abs/2512.24327v2

Spatial graphs are particular graphs for which the nodes are localized in space (e.g., public transport network, molecules, branching biological structures). In this work, we consider the problem of spatial graph reduction, that aims to find a smaller spatial graph (i.e., with less nodes) with the same overall structure as the initial one. In this context, performing the graph reduction while preserving the main topological features of the initial graph is particularly relevant, due to the additional spatial information. Thus, we propose a topological spatial graph coarsening approach based on a new framework that finds a trade-off between the graph reduction and the preservation of the topological characteristics. The coarsening is realized by collapsing short edges. In order to capture the topological information required to calibrate the reduction level, we adapt the construction of classical topological descriptors made for point clouds (the so-called persistent diagrams) to spatial graphs. This construction relies on the introduction of a new filtration called triangle-aware graph filtration. Our coarsening approach is parameter-free and we prove that it is equivariant under rotations, translations and scaling of the initial spatial graph. We evaluate the performances of our method on synthetic and real spatial graphs, and show that it significantly reduces the graph sizes while preserving the relevant topological information.

---

### MaRCA: Multi-Agent Reinforcement Learning for Dynamic Computation Allocation in Large-Scale Recommender Systems

Authors: Wan Jiang, Xinyi Zang, Yudong Zhao, Yusi Zou, Yunfei Lu, Junbo Tong, Yang Liu, Ming Li, Jiani Shi, Xin Yang
Published: 2025-12-30
Categories: cs.IR, cs.LG, cs.MA
arXiv: https://arxiv.org/abs/2512.24325v1

Modern recommender systems face significant computational challenges due to growing model complexity and traffic scale, making efficient computation allocation critical for maximizing business revenue. Existing approaches typically simplify multi-stage computation resource allocation, neglecting inter-stage dependencies, thus limiting global optimality. In this paper, we propose MaRCA, a multi-agent reinforcement learning framework for end-to-end computation resource allocation in large-scale recommender systems. MaRCA models the stages of a recommender system as cooperative agents, using Centralized Training with Decentralized Execution (CTDE) to optimize revenue under computation resource constraints. We introduce an AutoBucket TestBench for accurate computation cost estimation, and a Model Predictive Control (MPC)-based Revenue-Cost Balancer to proactively forecast traffic loads and adjust the revenue-cost trade-off accordingly. Since its end-to-end deployment in the advertising pipeline of a leading global e-commerce platform in November 2024, MaRCA has consistently handled hundreds of billions of ad requests per day and has delivered a 16.67% revenue uplift using existing computation resources.

---

### Empower Low-Altitude Economy: A Reliability-Aware Dynamic Weighting Allocation for Multi-modal UAV Beam Prediction

Authors: Haojin Li, Anbang Zhang, Chen Sun, Chenyuan Feng, Kaiqian Qu, Tony Q. S. Quek, Haijun Zhang
Published: 2025-12-30
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.24324v1

The low-altitude economy (LAE) is rapidly expanding driven by urban air mobility, logistics drones, and aerial sensing, while fast and accurate beam prediction in uncrewed aerial vehicles (UAVs) communications is crucial for achieving reliable connectivity. Current research is shifting from single-signal to multi-modal collaborative approaches. However, existing multi-modal methods mostly employ fixed or empirical weights, assuming equal reliability across modalities at any given moment. Indeed, the importance of different modalities fluctuates dramatically with UAV motion scenarios, and static weighting amplifies the negative impact of degraded modalities. Furthermore, modal mismatch and weak alignment further undermine cross-scenario generalization. To this end, we propose a reliability-aware dynamic weighting scheme applied to a semantic-aware multi-modal beam prediction framework, named SaM2B. Specifically, SaM2B leverages lightweight cues such as environmental visual, flight posture, and geospatial data to adaptively allocate contributions across modalities at different time points through reliability-aware dynamic weight updates. Moreover, by utilizing cross-modal contrastive learning, we align the "multi-source representation beam semantics" associated with specific beam information to a shared semantic space, thereby enhancing discriminative power and robustness under modal noise and distribution shifts. Experiments on real-world low-altitude UAV datasets show that SaM2B achieves more satisfactory results than baseline methods.

---

### Robust Egocentric Referring Video Object Segmentation via Dual-Modal Causal Intervention

Authors: Haijing Liu, Zhiyuan Song, Hefeng Wu, Tao Pu, Keze Wang, Liang Lin
Published: 2025-12-30
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.24323v1

Egocentric Referring Video Object Segmentation (Ego-RVOS) aims to segment the specific object actively involved in a human action, as described by a language query, within first-person videos. This task is critical for understanding egocentric human behavior. However, achieving such segmentation robustly is challenging due to ambiguities inherent in egocentric videos and biases present in training data. Consequently, existing methods often struggle, learning spurious correlations from skewed object-action pairings in datasets and fundamental visual confounding factors of the egocentric perspective, such as rapid motion and frequent occlusions. To address these limitations, we introduce Causal Ego-REferring Segmentation (CERES), a plug-in causal framework that adapts strong, pre-trained RVOS backbones to the egocentric domain. CERES implements dual-modal causal intervention: applying backdoor adjustment principles to counteract language representation biases learned from dataset statistics, and leveraging front-door adjustment concepts to address visual confounding by intelligently integrating semantic visual features with geometric depth information guided by causal principles, creating representations more robust to egocentric distortions. Extensive experiments demonstrate that CERES achieves state-of-the-art performance on Ego-RVOS benchmarks, highlighting the potential of applying causal reasoning to build more reliable models for broader egocentric video understanding.

---

### UniAct: Unified Motion Generation and Action Streaming for Humanoid Robots

Authors: Nan Jiang, Zimo He, Wanhe Yu, Lexi Pang, Yunhao Li, Hongjie Li, Jieming Cui, Yuhan Li, Yizhou Wang, Yixin Zhu, Siyuan Huang
Published: 2025-12-30
Categories: cs.CV, cs.RO
arXiv: https://arxiv.org/abs/2512.24321v1

A long-standing objective in humanoid robotics is the realization of versatile agents capable of following diverse multimodal instructions with human-level flexibility. Despite advances in humanoid control, bridging high-level multimodal perception with whole-body execution remains a significant bottleneck. Existing methods often struggle to translate heterogeneous instructions -- such as language, music, and trajectories -- into stable, real-time actions. Here we show that UniAct, a two-stage framework integrating a fine-tuned MLLM with a causal streaming pipeline, enables humanoid robots to execute multimodal instructions with sub-500 ms latency. By unifying inputs through a shared discrete codebook via FSQ, UniAct ensures cross-modal alignment while constraining motions to a physically grounded manifold. This approach yields a 19% improvement in the success rate of zero-shot tracking of imperfect reference motions. We validate UniAct on UniMoCap, our 20-hour humanoid motion benchmark, demonstrating robust generalization across diverse real-world scenarios. Our results mark a critical step toward responsive, general-purpose humanoid assistants capable of seamless interaction through unified perception and control.

---

### QianfanHuijin Technical Report: A Novel Multi-Stage Training Paradigm for Finance Industrial LLMs

Authors: Shupeng Li, Weipeng Lu, Linyun Liu, Chen Lin, Shaofei Li, Zhendong Tan, Hanjun Zhong, Yucheng Zeng, Chenghao Zhu, Mengyue Liu, Daxiang Dong, Jianmin Wu et al.
Published: 2025-12-30
Categories: cs.CL
arXiv: https://arxiv.org/abs/2512.24314v2

Domain-specific enhancement of Large Language Models (LLMs) within the financial context has long been a focal point of industrial application. While previous models such as BloombergGPT and Baichuan-Finance primarily focused on knowledge enhancement, the deepening complexity of financial services has driven a growing demand for models that possess not only domain knowledge but also robust financial reasoning and agentic capabilities. In this paper, we present QianfanHuijin, a financial domain LLM, and propose a generalizable multi-stage training paradigm for industrial model enhancement. Our approach begins with Continual Pre-training (CPT) on financial corpora to consolidate the knowledge base. This is followed by a fine-grained Post-training pipeline designed with increasing specificity: starting with Financial SFT, progressing to Finance Reasoning RL and Finance Agentic RL, and culminating in General RL aligned with real-world business scenarios. Empirical results demonstrate that QianfanHuijin achieves superior performance across various authoritative financial benchmarks. Furthermore, ablation studies confirm that the targeted Reasoning RL and Agentic RL stages yield significant gains in their respective capabilities. These findings validate our motivation and suggest that this fine-grained, progressive post-training methodology is poised to become a mainstream paradigm for various industrial-enhanced LLMs.

---

### Generative Video Compression: Towards 0.01% Compression Rate for Video Transmission

Authors: Xiangyu Chen, Jixiang Luo, Jingyu Xu, Fangqiu Yi, Chi Zhang, Xuelong Li
Published: 2025-12-30
Categories: eess.IV, cs.AI, cs.MM
arXiv: https://arxiv.org/abs/2512.24300v2

Whether a video can be compressed at an extreme compression rate as low as 0.01%? To this end, we achieve the compression rate as 0.02% at some cases by introducing Generative Video Compression (GVC), a new framework that redefines the limits of video compression by leveraging modern generative video models to achieve extreme compression rates while preserving a perception-centric, task-oriented communication paradigm, corresponding to Level C of the Shannon-Weaver model. Besides, How we trade computation for compression rate or bandwidth? GVC answers this question by shifting the burden from transmission to inference: it encodes video into extremely compact representations and delegates content reconstruction to the receiver, where powerful generative priors synthesize high-quality video from minimal transmitted information. Is GVC practical and deployable? To ensure practical deployment, we propose a compression-computation trade-off strategy, enabling fast inference on consume-grade GPUs. Within the AI Flow framework, GVC opens new possibility for video communication in bandwidth- and resource-constrained environments such as emergency rescue, remote surveillance, and mobile edge computing. Through empirical validation, we demonstrate that GVC offers a viable path toward a new effective, efficient, scalable, and practical video communication paradigm.

---

### Figure It Out: Improve the Frontier of Reasoning with Executable Visual States

Authors: Meiqi Chen, Fandong Meng, Jie Zhou
Published: 2025-12-30
Categories: cs.CL
arXiv: https://arxiv.org/abs/2512.24297v2

Complex reasoning problems often involve implicit spatial and geometric relationships that are not explicitly encoded in text. While recent reasoning models perform well across many domains, purely text-based reasoning struggles to capture structural constraints in complex settings. In this paper, we introduce FIGR, which integrates executable visual construction into multi-turn reasoning via end-to-end reinforcement learning. Rather than relying solely on textual chains of thought, FIGR externalizes intermediate hypotheses by generating executable code that constructs diagrams within the reasoning loop. An adaptive reward mechanism selectively regulates when visual construction is invoked, enabling more consistent reasoning over latent global properties that are difficult to infer from text alone. Experiments on eight challenging mathematical benchmarks demonstrate that FIGR outperforms strong text-only chain-of-thought baselines, improving the base model by 13.12% on AIME 2025 and 11.00% on BeyondAIME. These results highlight the effectiveness of precise, controllable figure construction of FIGR in enhancing complex reasoning ability.

---

### Virtual-Eyes: Quantitative Validation of a Lung CT Quality-Control Pipeline for Foundation-Model Cancer Risk Prediction

Authors: Md. Enamul Hoq, Linda Larson-Prior, Fred Prior
Published: 2025-12-30
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2512.24294v1

Robust preprocessing is rarely quantified in deep-learning pipelines for low-dose CT (LDCT) lung cancer screening. We develop and validate Virtual-Eyes, a clinically motivated 16-bit CT quality-control pipeline, and measure its differential impact on generalist foundation models versus specialist models. Virtual-Eyes enforces strict 512x512 in-plane resolution, rejects short or non-diagnostic series, and extracts a contiguous lung block using Hounsfield-unit filtering and bilateral lung-coverage scoring while preserving the native 16-bit grid. Using 765 NLST patients (182 cancer, 583 non-cancer), we compute slice-level embeddings from RAD-DINO and Merlin with frozen encoders and train leakage-free patient-level MLP heads; we also evaluate Sybil and a 2D ResNet-18 baseline under Raw versus Virtual-Eyes inputs without backbone retraining. Virtual-Eyes improves RAD-DINO slice-level AUC from 0.576 to 0.610 and patient-level AUC from 0.646 to 0.683 (mean pooling) and from 0.619 to 0.735 (max pooling), with improved calibration (Brier score 0.188 to 0.112). In contrast, Sybil and ResNet-18 degrade under Virtual-Eyes (Sybil AUC 0.886 to 0.837; ResNet-18 AUC 0.571 to 0.596) with evidence of context dependence and shortcut learning, and Merlin shows limited transferability (AUC approximately 0.507 to 0.567) regardless of preprocessing. These results demonstrate that anatomically targeted QC can stabilize and improve generalist foundation-model workflows but may disrupt specialist models adapted to raw clinical context.

---

### Fast reconstruction-based ROI triggering via anomaly detection in the CYGNO optical TPC

Authors: F. D. Amaro, R. Antonietti, E. Baracchini, L. Benussi, C. Capoccia, M. Caponero, L. G. M. de Carvalho, G. Cavoto, I. A. Costa, A. Croce, M. D'Astolfo, G. D'Imperio et al.
Published: 2025-12-30
Categories: physics.ins-det, cs.LG, physics.data-an
arXiv: https://arxiv.org/abs/2512.24290v2

Optical-readout Time Projection Chambers (TPCs) produce megapixel-scale images whose fine-grained topological information is essential for rare-event searches, but whose size challenges real-time data selection. We present an unsupervised, reconstruction-based anomaly-detection strategy for fast Region-of-Interest (ROI) extraction that operates directly on minimally processed camera frames. A convolutional autoencoder trained exclusively on pedestal images learns the detector noise morphology without labels, simulation, or fine-grained calibration. Applied to standard data-taking frames, localized reconstruction residuals identify particle-induced structures, from which compact ROIs are extracted via thresholding and spatial clustering. Using real data from the CYGNO optical TPC prototype, we compare two pedestal-trained autoencoder configurations that differ only in their training objective, enabling a controlled study of its impact. The best configuration retains (93.0 +/- 0.2)% of reconstructed signal intensity while discarding (97.8 +/- 0.1)% of the image area, with an inference time of approximately 25 ms per frame on a consumer GPU. The results demonstrate that careful design of the training objective is critical for effective reconstruction-based anomaly detection and that pedestal-trained autoencoders provide a transparent and detector-agnostic baseline for online data reduction in optical TPCs.

---

### Automated Analysis of Sustainability Reports: Using Large Language Models for the Extraction and Prediction of EU Taxonomy-Compliant KPIs

Authors: Jonathan Schmoll, Adam Jatowt
Published: 2025-12-30
Categories: cs.CL
arXiv: https://arxiv.org/abs/2512.24289v1

The manual, resource-intensive process of complying with the EU Taxonomy presents a significant challenge for companies. While Large Language Models (LLMs) offer a path to automation, research is hindered by a lack of public benchmark datasets. To address this gap, we introduce a novel, structured dataset from 190 corporate reports, containing ground-truth economic activities and quantitative Key Performance Indicators (KPIs). We use this dataset to conduct the first systematic evaluation of LLMs on the core compliance workflow. Our results reveal a clear performance gap between qualitative and quantitative tasks. LLMs show moderate success in the qualitative task of identifying economic activities, with a multi-step agentic framework modestly enhancing precision. Conversely, the models comprehensively fail at the quantitative task of predicting financial KPIs in a zero-shot setting. We also discover a paradox, where concise metadata often yields superior performance to full, unstructured reports, and find that model confidence scores are poorly calibrated. We conclude that while LLMs are not ready for full automation, they can serve as powerful assistive tools for human experts. Our dataset provides a public benchmark for future research.

---