# Research Papers — 2025-12 · bundle page 138 of 175

> Full-text markdown of papers #6851–6900 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/138
[← Previous bundle](https://aipod.dev/papers/archive/2025-12/page/137.md)
[Next bundle →](https://aipod.dev/papers/archive/2025-12/page/139.md)

### SPOOF: Simple Pixel Operations for Out-of-Distribution Fooling

Authors: Ankit Gupta, Christoph Adami, Emily Dolson
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.06185v1

Deep neural networks (DNNs) excel across image recognition tasks, yet continue to exhibit overconfidence on inputs that bear no resemblance to natural images. Revisiting the "fooling images" work introduced by Nguyen et al. (2015), we re-implement both CPPN-based and direct-encoding-based evolutionary fooling attacks on modern architectures, including convolutional and transformer classifiers. Our re-implementation confirm that high-confidence fooling persists even in state-of-the-art networks, with transformer-based ViT-B/16 emerging as the most susceptible--achieving near-certain misclassifications with substantially fewer queries than convolution-based models. We then introduce SPOOF, a minimalist, consistent, and more efficient black-box attack generating high-confidence fooling images. Despite its simplicity, SPOOF generates unrecognizable fooling images with minimal pixel modifications and drastically reduced compute. Furthermore, retraining with fooling images as an additional class provides only partial resistance, as SPOOF continues to fool consistently with slightly higher query budgets--highlighting persistent fragility of modern deep classifiers.

---

### PMA-Diffusion: A Physics-guided Mask-Aware Diffusion Framework for TSE from Sparse Observations

Authors: Lindong Liu, Zhixiong Jin, Seongjin Choi
Published: 2025-12-05
Categories: cs.LG
arXiv: https://arxiv.org/abs/2512.06183v1

High-resolution highway traffic state information is essential for Intelligent Transportation Systems, but typical traffic data acquired from loop detectors and probe vehicles are often too sparse and noisy to capture the detailed dynamics of traffic flow. We propose PMA-Diffusion, a physics-guided mask-aware diffusion framework that reconstructs unobserved highway speed fields from sparse, incomplete observations. Our approach trains a diffusion prior directly on sparsely observed speed fields using two mask-aware training strategies: Single-Mask and Double-Mask. At the inference phase, the physics-guided posterior sampler alternates reverse-diffusion updates, observation projection, and physics-guided projection based on adaptive anisotropic smoothing to reconstruct the missing speed fields. The proposed framework is tested on the I-24 MOTION dataset with varying visibility ratios. Even under severe sparsity, with only 5% visibility, PMA-Diffusion outperforms other baselines across three reconstruction error metrics. Furthermore, PMA-diffusion trained with sparse observation nearly matches the performance of the baseline model trained on fully observed speed fields. The results indicate that combining mask-aware diffusion priors with a physics-guided posterior sampler provides a reliable and flexible solution for traffic state estimation under realistic sensing sparsity.

---

### Beyond Lux thresholds: a systematic pipeline for classifying biologically relevant light contexts from wearable data

Authors: Yanuo Zhou
Published: 2025-12-05
Categories: q-bio.QM, cs.LG
arXiv: https://arxiv.org/abs/2512.06181v2

Background: Wearable spectrometers enable field quantification of biologically relevant light, yet reproducible pipelines for contextual classification remain under-specified. Objective: To establish and validate a subject-wise evaluated, reproducible pipeline and actionable design rules for classifying natural vs. artificial light from wearable spectral data. Methods: We analysed ActLumus recordings from 26 participants, each monitored for at least 7 days at 10-second sampling, paired with daily exposure diaries. The pipeline fixes the sequence: domain selection, log-base-10 transform, L2 normalisation excluding total intensity (to avoid brightness shortcuts), hour-level medoid aggregation, sine/cosine hour encoding, and MLP classifier, evaluated under participant-wise cross-validation. Results: The proposed sequence consistently achieved high performance on the primary task, with representative configurations reaching AUC = 0.938 (accuracy 88%) for natural vs. artificial classification on the held-out subject split. In contrast, indoor vs. outdoor classification remained at feasibility level due to spectral overlap and class imbalance (best AUC approximately 0.75; majority-class collapse without contextual sensors). Threshold baselines were insufficient on our data, supporting the need for spectral-temporal modelling beyond illuminance cut-offs. Conclusions: We provide a reproducible, auditable baseline pipeline and design rules for contextual light classification under subject-wise generalisation. All code, configuration files, and derived artefacts will be openly archived (GitHub + Zenodo DOI) to support reuse and benchmarking.

---

### Cast and Attached Shadow Detection via Iterative Light and Geometry Reasoning

Authors: Shilin Hu, Jingyi Xu, Sagnik Das, Dimitris Samaras, Hieu Le
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.06179v3

Shadows encode rich information about scene geometry and illumination, yet existing methods either predict a unified shadow mask or overlook attached shadows entirely. We address this gap by proposing a framework for jointly detecting cast and attached shadows through explicit physical modeling of light direction and surface geometry under a dominant directional-light setting. Our approach is grounded in a simple observation: surfaces facing away from the light source tend to fall into shadow. We exploit the reciprocal relationship between shadow formation and light estimation to construct a closed feedback loop, a dual-module architecture in which a shadow detection module and a light estimation module iteratively refine each other. At each pass, updated light estimates, together with surface normals, produce partial attached shadow maps that guide detection, while improved shadow predictions sharpen light estimation. To support training and evaluation, we introduce a dataset of 1,458 images with manually annotated cast and attached shadow masks sourced from three existing benchmarks. Experiments demonstrate that our proposed method outperforms prior methods, with at least a 33% reduction in attached-shadow BER, while maintaining strong full-shadow and cast-shadow performance.

---

### Embedding Physical Reasoning into Diffusion-Based Shadow Generation Under the Sun and Sky

Authors: Shilin Hu, Jingyi Xu, Akshat Dave, Dimitris Samaras, Hieu Le
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.06174v3

Generating realistic cast shadows for inserted foreground objects requires reasoning about scene geometry and illumination. However, most learning-based approaches treat shadow generation as an image translation problem and capture these physical relationships only implicitly. This often results in misaligned or implausible shadows. Motivated by the physics of shadow formation, we introduce explicit geometric guidance for outdoor shadow generation. Given a composite image and a foreground object mask, we recover approximate scene geometry and estimate a dominant light direction to derive a coarse shadow estimate via geometric reasoning. While coarse, this estimate provides a spatial anchor for shadow placement. Because illumination cannot always be uniquely inferred from a single image, we predict confidence scores for both lighting and shadow cues and use them to regulate their influence during generation. These cues (shadow mask, light direction, and their confidence scores) condition a diffusion-based generator that refines the estimate into a realistic shadow. Experiments on DESOBAV2 show substantially improved shadow-region fidelity and localization, with an overall 23% lower shadow-region RMSE and 30% lower shadow-mask BER than the prior state-of-the-art method.

---

### DEFEND: Poisoned Model Detection and Malicious Client Exclusion Mechanism for Secure Federated Learning-based Road Condition Classification

Authors: Sheng Liu, Panos Papadimitratos
Published: 2025-12-05
Categories: cs.CR, cs.AI
arXiv: https://arxiv.org/abs/2512.06172v1

Federated Learning (FL) has drawn the attention of the Intelligent Transportation Systems (ITS) community. FL can train various models for ITS tasks, notably camera-based Road Condition Classification (RCC), in a privacy-preserving collaborative way. However, opening up to collaboration also opens FL-based RCC systems to adversaries, i.e., misbehaving participants that can launch Targeted Label-Flipping Attacks (TLFAs) and threaten transportation safety. Adversaries mounting TLFAs poison training data to misguide model predictions, from an actual source class (e.g., wet road) to a wrongly perceived target class (e.g., dry road). Existing countermeasures against poisoning attacks cannot maintain model performance under TLFAs close to the performance level in attack-free scenarios, because they lack specific model misbehavior detection for TLFAs and neglect client exclusion after the detection. To close this research gap, we propose DEFEND, which includes a poisoned model detection strategy that leverages neuron-wise magnitude analysis for attack goal identification and Gaussian Mixture Model (GMM)-based clustering. DEFEND discards poisoned model contributions in each round and adapts accordingly client ratings, eventually excluding malicious clients. Extensive evaluation involving various FL-RCC models and tasks shows that DEFEND can thwart TLFAs and outperform seven baseline countermeasures, with at least 15.78% improvement, with DEFEND remarkably achieving under attack the same performance as in attack-free scenarios.

---

### Automated Annotation of Shearographic Measurements Enabling Weakly Supervised Defect Detection

Authors: Jessica Plassmann, Nicolas Schuler, Michael Schuth, Georg von Freymann
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.06171v2

Shearography is an interferometric technique sensitive to surface displacement gradients, providing high sensitivity for detecting subsurface defects in safety-critical components. A key limitation to industrial adoption is the lack of high-quality annotated datasets, since manual labeling remains labor-intensive, subjective, and difficult to standardize. We present an automated labeling pipeline that generates candidate defect bounding boxes with Grounded DINO, refines them using SAM masks, and exports YOLO-format labels for downstream detector training. Quantitative evaluation shows the generated boxes are suitable for weakly supervised learning, while high-resolution masks provide qualitative visualization. This approach reduces manual effort and supports scalable dataset creation for robust industrial defect detection.

---

### Morphologically-Informed Tokenizers for Languages with Non-Concatenative Morphology: A case study of Yoloxóchtil Mixtec ASR

Authors: Chris Crawford
Published: 2025-12-05
Categories: cs.CL
arXiv: https://arxiv.org/abs/2512.06169v1

This paper investigates the impact of using morphologically-informed tokenizers to aid and streamline the interlinear gloss annotation of an audio corpus of Yoloxóchitl Mixtec (YM) using a combination of ASR and text-based sequence-to-sequence tools, with the goal of improving efficiency while reducing the workload of a human annotator. We present two novel tokenization schemes that separate words in a nonlinear manner, preserving information about tonal morphology as much as possible. One of these approaches, a Segment and Melody tokenizer, simply extracts the tones without predicting segmentation. The other, a Sequence of Processes tokenizer, predicts segmentation for the words, which could allow an end-to-end ASR system to produce segmented and unsegmented transcriptions in a single pass. We find that these novel tokenizers are competitive with BPE and Unigram models, and the Segment-and-Melody model outperforms traditional tokenizers in terms of word error rate but does not reach the same character error rate. In addition, we analyze tokenizers on morphological and information-theoretic metrics to find predictive correlations with downstream performance. Our results suggest that nonlinear tokenizers designed specifically for the non-concatenative morphology of a language are competitive with conventional BPE and Unigram models for ASR. Further research will be necessary to determine the applicability of these tokenizers in downstream processing tasks.

---

### Deep learning for autism detection using clinical notes: A comparison of transfer learning for a transparent and black-box approach

Authors: Gondy Leroy, Prakash Bisht, Sai Madhuri Kandula, Nell Maltman, Sydney Rice
Published: 2025-12-05
Categories: cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2512.06161v1

Autism spectrum disorder (ASD) is a complex neurodevelopmental condition whose rising prevalence places increasing demands on a lengthy diagnostic process. Machine learning (ML) has shown promise in automating ASD diagnosis, but most existing models operate as black boxes and are typically trained on a single dataset, limiting their generalizability. In this study, we introduce a transparent and interpretable ML approach that leverages BioBERT, a state-of-the-art language model, to analyze unstructured clinical text. The model is trained to label descriptions of behaviors and map them to diagnostic criteria, which are then used to assign a final label (ASD or not). We evaluate transfer learning, the ability to transfer knowledge to new data, using two distinct real-world datasets. We trained on datasets sequentially and mixed together and compared the performance of the best models and their ability to transfer to new data. We also created a black-box approach and repeated this transfer process for comparison. Our transparent model demonstrated robust performance, with the mixed-data training strategy yielding the best results (97 % sensitivity, 98 % specificity). Sequential training across datasets led to a slight drop in performance, highlighting the importance of training data order. The black-box model performed worse (90 % sensitivity, 96 % specificity) when trained sequentially or with mixed data. Overall, our transparent approach outperformed the black-box approach. Mixing datasets during training resulted in slightly better performance and should be the preferred approach when practically possible. This work paves the way for more trustworthy, generalizable, and clinically actionable AI tools in neurodevelopmental diagnostics.

---

### Tracking-Guided 4D Generation: Foundation-Tracker Motion Priors for 3D Model Animation

Authors: Su Sun, Cheng Zhao, Himangi Mittal, Gaurav Mittal, Rohith Kukkala, Yingjie Victor Chen, Mei Chen
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.06158v1

Generating dynamic 4D objects from sparse inputs is difficult because it demands joint preservation of appearance and motion coherence across views and time while suppressing artifacts and temporal drift. We hypothesize that the view discrepancy arises from supervision limited to pixel- or latent-space video-diffusion losses, which lack explicitly temporally aware, feature-level tracking guidance. We present \emph{Track4DGen}, a two-stage framework that couples a multi-view video diffusion model with a foundation point tracker and a hybrid 4D Gaussian Splatting (4D-GS) reconstructor. The central idea is to explicitly inject tracker-derived motion priors into intermediate feature representations for both multi-view video generation and 4D-GS. In Stage One, we enforce dense, feature-level point correspondences inside the diffusion generator, producing temporally consistent features that curb appearance drift and enhance cross-view coherence. In Stage Two, we reconstruct a dynamic 4D-GS using a hybrid motion encoding that concatenates co-located diffusion features (carrying Stage-One tracking priors) with Hex-plane features, and augment them with 4D Spherical Harmonics for higher-fidelity dynamics modeling. \emph{Track4DGen} surpasses baselines on both multi-view video generation and 4D generation benchmarks, yielding temporally stable, text-editable 4D assets. Lastly, we curate \emph{Sketchfab28}, a high-quality dataset for benchmarking object-centric 4D generation and fostering future research.

---

### Learning Invariant Graph Representations Through Redundant Information

Authors: Barproda Halder, Pasan Dissanayake, Sanghamitra Dutta
Published: 2025-12-05
Categories: cs.LG, cs.AI, cs.IT
arXiv: https://arxiv.org/abs/2512.06154v1

Learning invariant graph representations for out-of-distribution (OOD) generalization remains challenging because the learned representations often retain spurious components. To address this challenge, this work introduces a new tool from information theory called Partial Information Decomposition (PID) that goes beyond classical information-theoretic measures. We identify limitations in existing approaches for invariant representation learning that solely rely on classical information-theoretic measures, motivating the need to precisely focus on redundant information about the target $Y$ shared between spurious subgraphs $G_s$ and invariant subgraphs $G_c$ obtained via PID. Next, we propose a new multi-level optimization framework that we call -- Redundancy-guided Invariant Graph learning (RIG) -- that maximizes redundant information while isolating spurious and causal subgraphs, enabling OOD generalization under diverse distribution shifts. Our approach relies on alternating between estimating a lower bound of redundant information (which itself requires an optimization) and maximizing it along with additional objectives. Experiments on both synthetic and real-world graph datasets demonstrate the generalization capabilities of our proposed RIG framework.

---

### GuideNav: User-Informed Development of a Vision-Only Robotic Navigation Assistant For Blind Travelers

Authors: Hochul Hwang, Soowan Yang, Jahir Sadik Monon, Nicholas A Giudice, Sunghoon Ivan Lee, Joydeep Biswas, Donghyun Kim
Published: 2025-12-05
Categories: cs.RO, cs.CV, cs.HC
arXiv: https://arxiv.org/abs/2512.06147v1

While commendable progress has been made in user-centric research on mobile assistive systems for blind and low-vision (BLV) individuals, references that directly inform robot navigation design remain rare. To bridge this gap, we conducted a comprehensive human study involving interviews with 26 guide dog handlers, four white cane users, nine guide dog trainers, and one O\&M trainer, along with 15+ hours of observing guide dog-assisted walking. After de-identification, we open-sourced the dataset to promote human-centered development and informed decision-making for assistive systems for BLV people. Building on insights from this formative study, we developed GuideNav, a vision-only, teach-and-repeat navigation system. Inspired by how guide dogs are trained and assist their handlers, GuideNav autonomously repeats a path demonstrated by a sighted person using a robot. Specifically, the system constructs a topological representation of the taught route, integrates visual place recognition with temporal filtering, and employs a relative pose estimator to compute navigation actions - all without relying on costly, heavy, power-hungry sensors such as LiDAR. In field tests, GuideNav consistently achieved kilometer-scale route following across five outdoor environments, maintaining reliability despite noticeable scene variations between teach and repeat runs. A user study with 3 guide dog handlers and 1 guide dog trainer further confirmed the system's feasibility, marking (to our knowledge) the first demonstration of a quadruped mobile system retrieving a path in a manner comparable to guide dogs.

---

### gp2Scale: A Class of Compactly Supported Non-Stationary Kernels and Distributed Computing for Exact Gaussian Processes on 10 Million Data Points

Authors: Marcus M. Noack, Mark D. Risser, Hengrui Luo, Vardaan Tekriwal, Ronald J. Pandolfi
Published: 2025-12-05
Categories: cs.LG, math.PR
arXiv: https://arxiv.org/abs/2512.06143v3

Despite a large corpus of recent work on scaling up Gaussian processes, a stubborn trade-off between computational speed, prediction and uncertainty quantification accuracy, and customizability persists. This is because the vast majority of existing methodologies exploit various levels of approximations that lower accuracy and limit the flexibility of kernel and noise-model designs -- an unacceptable drawback at a time when expressive non-stationary kernels are on the rise in many fields. Here, we propose a methodology we term \emph{gp2Scale} that scales exact Gaussian processes to more than 10 million data points without relying on inducing points, kernel interpolation, or neighborhood-based approximations, and instead leveraging the existing capabilities of a GP: its kernel design. Highly flexible, compactly supported, and non-stationary kernels lead to the identification of naturally occurring sparse structure in the covariance matrix, which is then exploited for the calculations of the linear system solution and the log-determinant for training. We demonstrate our method's functionality on several real-world datasets and compare it with state-of-the-art approximation algorithms. Although we show superior approximation performance in many cases, the method's real power lies in its agnosticism toward arbitrary GP customizations -- core kernel design, noise, and mean functions -- and the type of input space, making it optimally suited for modern Gaussian process applications.

---

### Entropic Regularization in the Deep Linear Network

Authors: Alan Chen, Tejas Kotwal, Govind Menon
Published: 2025-12-05
Categories: cs.NE, math.DS, math.PR
arXiv: https://arxiv.org/abs/2512.06137v1

We study regularization for the deep linear network (DLN) using the entropy formula introduced in arXiv:2509.09088. The equilibria and gradient flow of the free energy on the Riemannian manifold of end-to-end maps of the DLN are characterized for energies that depend symmetrically on the singular values of the end-to-end matrix. The only equilibria are minimizers and the set of minimizers is an orbit of the orthogonal group. In contrast with random matrix theory there is no singular value repulsion. The corresponding gradient flow reduces to a one-dimensional ordinary differential equation whose solution gives explicit relaxation rates toward the minimizers. We also study the concavity of the entropy in the chamber of singular values. The entropy is shown to be strictly concave in the Euclidean geometry on the chamber but not in the Riemannian geometry defined by the DLN metric.

---

### Physics-Informed Neural Koopman Machine for Interpretable Longitudinal Personalized Alzheimer's Disease Forecasting

Authors: Georgi Hrusanov, Duy-Thanh Vu, Duy-Cat Can, Sophie Tascedda, Margaret Ryan, Julien Bodelet, Katarzyna Koscielska, Carsten Magnus, Oliver Y. Chén
Published: 2025-12-05
Categories: cs.LG, cs.AI, q-bio.NC, q-bio.QM
arXiv: https://arxiv.org/abs/2512.06134v1

Early forecasting of individual cognitive decline in Alzheimer's disease (AD) is central to disease evaluation and management. Despite advances, it is as of yet challenging for existing methodological frameworks to integrate multimodal data for longitudinal personalized forecasting while maintaining interpretability. To address this gap, we present the Neural Koopman Machine (NKM), a new machine learning architecture inspired by dynamical systems and attention mechanisms, designed to forecast multiple cognitive scores simultaneously using multimodal genetic, neuroimaging, proteomic, and demographic data. NKM integrates analytical ($α$) and biological ($β$) knowledge to guide feature grouping and control the hierarchical attention mechanisms to extract relevant patterns. By implementing Fusion Group-Aware Hierarchical Attention within the Koopman operator framework, NKM transforms complex nonlinear trajectories into interpretable linear representations. To demonstrate NKM's efficacy, we applied it to study the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset. Our results suggest that NKM consistently outperforms both traditional machine learning methods and deep learning models in forecasting trajectories of cognitive decline. Specifically, NKM (1) forecasts changes of multiple cognitive scores simultaneously, (2) quantifies differential biomarker contributions to predicting distinctive cognitive scores, and (3) identifies brain regions most predictive of cognitive deterioration. Together, NKM advances personalized, interpretable forecasting of future cognitive decline in AD using past multimodal data through an explainable, explicit system and reveals potential multimodal biological underpinnings of AD progression.

---

### Toward Patch Robustness Certification and Detection for Deep Learning Systems Beyond Consistent Samples

Authors: Qilin Zhou, Zhengyuan Wei, Haipeng Wang, Zhuo Wang, W. K. Chan
Published: 2025-12-05
Categories: cs.SE, cs.AI, cs.CR
arXiv: https://arxiv.org/abs/2512.06123v1

Patch robustness certification is an emerging kind of provable defense technique against adversarial patch attacks for deep learning systems. Certified detection ensures the detection of all patched harmful versions of certified samples, which mitigates the failures of empirical defense techniques that could (easily) be compromised. However, existing certified detection methods are ineffective in certifying samples that are misclassified or whose mutants are inconsistently pre icted to different labels. This paper proposes HiCert, a novel masking-based certified detection technique. By focusing on the problem of mutants predicted with a label different from the true label with our formal analysis, HiCert formulates a novel formal relation between harmful samples generated by identified loopholes and their benign counterparts. By checking the bound of the maximum confidence among these potentially harmful (i.e., inconsistent) mutants of each benign sample, HiCert ensures that each harmful sample either has the minimum confidence among mutants that are predicted the same as the harmful sample itself below this bound, or has at least one mutant predicted with a label different from the harmful sample itself, formulated after two novel insights. As such, HiCert systematically certifies those inconsistent samples and consistent samples to a large extent. To our knowledge, HiCert is the first work capable of providing such a comprehensive patch robustness certification for certified detection. Our experiments show the high effectiveness of HiCert with a new state-of the-art performance: It certifies significantly more benign samples, including those inconsistent and consistent, and achieves significantly higher accuracy on those samples without warnings and a significantly lower false silent ratio.

---

### Hardware Software Optimizations for Fast Model Recovery on Reconfigurable Architectures

Authors: Bin Xu, Ayan Banerjee, Sandeep Gupta
Published: 2025-12-05
Categories: cs.AR, cs.LG
arXiv: https://arxiv.org/abs/2512.06113v1

Model Recovery (MR) is a core primitive for physical AI and real-time digital twins, but GPUs often execute MR inefficiently due to iterative dependencies, kernel-launch overheads, underutilized memory bandwidth, and high data-movement latency. We present MERINDA, an FPGA-accelerated MR framework that restructures computation as a streaming dataflow pipeline. MERINDA exploits on-chip locality through BRAM tiling, fixed-point kernels, and the concurrent use of LUT fabric and carry-chain adders to expose fine-grained spatial parallelism while minimizing off-chip traffic. This hardware-aware formulation removes synchronization bottlenecks and sustains high throughput across the iterative updates in MR. On representative MR workloads, MERINDA delivers up to 6.3x fewer cycles than an FPGA-based LTC baseline, enabling real-time performance for time-critical physical systems.

---

### WAM-Flow: Parallel Coarse-to-Fine Motion Planning via Discrete Flow Matching for Autonomous Driving

Authors: Yifang Xu, Jiahao Cui, Feipeng Cai, Zhihao Zhu, Hanlin Shang, Shan Luan, Mingwang Xu, Neng Zhang, Yaoyi Li, Jia Cai, Siyu Zhu
Published: 2025-12-05
Categories: cs.RO, cs.AI, cs.CV
arXiv: https://arxiv.org/abs/2512.06112v2

We introduce WAM-Flow, a vision-language-action (VLA) model that casts ego-trajectory planning as discrete flow matching over a structured token space. In contrast to autoregressive decoders, WAM-Flow performs fully parallel, bidirectional denoising, enabling coarse-to-fine refinement with a tunable compute-accuracy trade-off. Specifically, the approach combines a metric-aligned numerical tokenizer that preserves scalar geometry via triplet-margin learning, a geometry-aware flow objective and a simulator-guided GRPO alignment that integrates safety, ego progress, and comfort rewards while retaining parallel generation. A multi-stage adaptation converts a pre-trained auto-regressive backbone (Janus-1.5B) from causal decoding to non-causal flow model and strengthens road-scene competence through continued multimodal pretraining. Thanks to the inherent nature of consistency model training and parallel decoding inference, WAM-Flow achieves superior closed-loop performance against autoregressive and diffusion-based VLA baselines, with 1-step inference attaining 89.1 PDMS and 5-step inference reaching 90.3 PDMS on NAVSIM v1 benchmark. These results establish discrete flow matching as a new promising paradigm for end-to-end autonomous driving. The code will be publicly available soon.

---

### A Prescriptive Framework for Determining Optimal Days for Short-Term Traffic Counts

Authors: Arthur Mukwaya, Nancy Kasamala, Nana Kankam Gyimah, Judith Mwakalonge, Gurcan Comert, Saidi Siuhi, Denis Ruganuza, Mark Ngotonie
Published: 2025-12-05
Categories: cs.LG
arXiv: https://arxiv.org/abs/2512.06111v1

The Federal Highway Administration (FHWA) mandates that state Departments of Transportation (DOTs) collect reliable Annual Average Daily Traffic (AADT) data. However, many U.S. DOTs struggle to obtain accurate AADT, especially for unmonitored roads. While continuous count (CC) stations offer accurate traffic volume data, their implementation is expensive and difficult to deploy widely, compelling agencies to rely on short-duration traffic counts. This study proposes a machine learning framework, the first to our knowledge, to identify optimal representative days for conducting short count (SC) data collection to improve AADT prediction accuracy. Using 2022 and 2023 traffic volume data from the state of Texas, we compare two scenarios: an 'optimal day' approach that iteratively selects the most informative days for AADT estimation and a 'no optimal day' baseline reflecting current practice by most DOTs. To align with Texas DOT's traffic monitoring program, continuous count data were utilized to simulate the 24 hour short counts. The actual field short counts were used to enhance feature engineering through using a leave-one-out (LOO) technique to generate unbiased representative daily traffic features across similar road segments. Our proposed methodology outperforms the baseline across the top five days, with the best day (Day 186) achieving lower errors (RMSE: 7,871.15, MAE: 3,645.09, MAPE: 11.95%) and higher R^2 (0.9756) than the baseline (RMSE: 11,185.00, MAE: 5,118.57, MAPE: 14.42%, R^2: 0.9499). This research offers DOTs an alternative to conventional short-duration count practices, improving AADT estimation, supporting Highway Performance Monitoring System compliance, and reducing the operational costs of statewide traffic data collection.

---

### Unifying Entropy Regularization in Optimal Control: From and Back to Classical Objectives via Iterated Soft Policies and Path Integral Solutions

Authors: Ajinkya Bhole, Mohammad Mahmoudi Filabadi, Guillaume Crevecoeur, Tom Lefebvre
Published: 2025-12-05
Categories: math.OC, cs.LG, cs.RO, eess.SY
arXiv: https://arxiv.org/abs/2512.06109v3

This paper develops a unified perspective on several optimal control formulations through the lens of Kullback-Leibler (KL) regularization. We propose a central problem that separates the KL penalties on policies and transitions with independent weights, thus generalizing the standard trajectory-level KL-regularization used in probabilistic optimal control. This umbrella formulation recovers various control problems: the classical Stochastic Optimal Control (SOC), Risk-Sensitive Stochastic Optimal Control (RSOC), and their policy-based KL-regularized counterparts, termed soft-policy SOC and RSOC, which yield tractable surrogates. Beyond being regularized variants, these soft-policy formulations majorize the original SOC and RSOC, thus, iterating their solutions recovers the original objectives. We further identify a synchronized case of soft-policy RSOC where the policy and transition KL weights coincide, yielding a linear Bellman operator, path-integral solution, and compositionality -- extending these computationally favourable properties to a broad class of control problems.

---

### From Tail Universality to Bernstein-von Mises: A Unified Statistical Theory of Semi-Implicit Variational Inference

Authors: Sean Plummer
Published: 2025-12-05
Categories: math.ST, stat.ML
arXiv: https://arxiv.org/abs/2512.06107v1

Semi-implicit variational inference (SIVI) constructs approximate posteriors of the form $q(θ) = \int k(θ| z) r(dz)$, where the conditional kernel is parameterized and the mixing base is fixed and tractable. This paper develops a unified "approximation-optimization-statistics'' theory for such families. On the approximation side, we show that under compact L1-universality and a mild tail-dominance condition, semi-implicit families are dense in L1 and can achieve arbitrarily small forward Kullback-Leibler (KL) error. We also identify two sharp obstructions to global approximation: (i) an Orlicz tail-mismatch condition that induces a strictly positive forward-KL gap, and (ii) structural restrictions, such as non-autoregressive Gaussian kernels, that force "branch collapse'' in conditional distributions. For each obstruction we give a minimal structural modification that restores approximability. On the optimization side, we establish finite-sample oracle inequalities and prove that the empirical SIVI objectives L(K,n) $Γ$-converge to their population limit as n and K tend to infinity. These results give consistency of empirical maximizers, quantitative control of finite-K surrogate bias, and stability of the resulting variational posteriors. Combining the approximation and optimization analyses yields the first general end-to-end statistical theory for SIVI: we characterize precisely when SIVI can recover the target distribution, when it cannot, and how architectural and algorithmic choices govern the attainable asymptotic behavior.

---

### Future You: Designing and Evaluating Multimodal AI-generated Digital Twins for Strengthening Future Self-Continuity

Authors: Constanze Albrecht, Chayapatr Archiwaranguprok, Rachel Poonsiriwong, Awu Chen, Peggy Yin, Monchai Lertsutthiwong, Kavin Winson, Hal Hershfield, Pattie Maes, Pat Pataranutaporn
Published: 2025-12-05
Categories: cs.HC, cs.AI
arXiv: https://arxiv.org/abs/2512.06106v1

What if users could meet their future selves today? AI-generated future selves simulate meaningful encounters with a digital twin decades in the future. As AI systems advance, combining cloned voices, age-progressed facial rendering, and autobiographical narratives, a central question emerges: Does the modality of these future selves alter their psychological and affective impact? How might a text-based chatbot, a voice-only system, or a photorealistic avatar shape present-day decisions and our feeling of connection to the future? We report a randomized controlled study (N=92) evaluating three modalities of AI-generated future selves (text, voice, avatar) against a neutral control condition. We also report a systematic model evaluation between Claude 4 and three other Large Language Models (LLMs), assessing Claude 4 across psychological and interaction dimensions and establishing conversational AI quality as a critical determinant of intervention effectiveness. All personalized modalities strengthened Future Self-Continuity (FSC), emotional well-being, and motivation compared to control, with avatar producing the largest vividness gains, yet with no significant differences between formats. Interaction quality metrics, particularly persuasiveness, realism, and user engagement, emerged as robust predictors of psychological and affective outcomes, indicating that how compelling the interaction feels matters more than the form it takes. Content analysis found thematic patterns: text emphasized career planning, while voice and avatar facilitated personal reflection. Claude 4 outperformed ChatGPT 3.5, Llama 4, and Qwen 3 in enhancing psychological, affective, and FSC outcomes.

---

### Explainable Melanoma Diagnosis with Contrastive Learning and LLM-based Report Generation

Authors: Junwen Zheng, Xinran Xu, Li Rong Wang, Chang Cai, Lucinda Siyun Tan, Dingyuan Wang, Hong Liang Tey, Xiuyi Fan
Published: 2025-12-05
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2512.06105v1

Deep learning has demonstrated expert-level performance in melanoma classification, positioning it as a powerful tool in clinical dermatology. However, model opacity and the lack of interpretability remain critical barriers to clinical adoption, as clinicians often struggle to trust the decision-making processes of black-box models. To address this gap, we present a Cross-modal Explainable Framework for Melanoma (CEFM) that leverages contrastive learning as the core mechanism for achieving interpretability. Specifically, CEFM maps clinical criteria for melanoma diagnosis-namely Asymmetry, Border, and Color (ABC)-into the Vision Transformer embedding space using dual projection heads, thereby aligning clinical semantics with visual features. The aligned representations are subsequently translated into structured textual explanations via natural language generation, creating a transparent link between raw image data and clinical interpretation. Experiments on public datasets demonstrate 92.79% accuracy and an AUC of 0.961, along with significant improvements across multiple interpretability metrics. Qualitative analyses further show that the spatial arrangement of the learned embeddings aligns with clinicians' application of the ABC rule, effectively bridging the gap between high-performance classification and clinical trust.

---

### ARC-AGI Without Pretraining

Authors: Isaac Liao, Albert Gu
Published: 2025-12-05
Categories: cs.LG
arXiv: https://arxiv.org/abs/2512.06104v1

Conventional wisdom in the age of LLMs dictates that solving IQ-test-like visual puzzles from the ARC-AGI-1 benchmark requires capabilities derived from massive pretraining. To counter this, we introduce CompressARC, a 76K parameter model without any pretraining that solves 20% of evaluation puzzles by minimizing the description length (MDL) of the target puzzle purely during inference time. The MDL endows CompressARC with extreme generalization abilities typically unheard of in deep learning. To our knowledge, CompressARC is the only deep learning method for ARC-AGI where training happens only on a single sample: the target inference puzzle itself, with the final solution information removed. Moreover, CompressARC does not train on the pre-provided ARC-AGI "training set". Under these extremely data-limited conditions, we do not ordinarily expect any puzzles to be solvable at all. Yet CompressARC still solves a diverse distribution of creative ARC-AGI puzzles, suggesting MDL to be an alternative feasible way to produce intelligence, besides conventional pretraining.

---

### SpectraIrisPAD: Leveraging Vision Foundation Models for Spectrally Conditioned Multispectral Iris Presentation Attack Detection

Authors: Raghavendra Ramachandra, Sushma Venkatesh
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.06103v1

Iris recognition is widely recognized as one of the most accurate biometric modalities. However, its growing deployment in real-world applications raises significant concerns regarding its vulnerability to Presentation Attacks (PAs). Effective Presentation Attack Detection (PAD) is therefore critical to ensure the integrity and security of iris-based biometric systems. While conventional iris recognition systems predominantly operate in the near-infrared (NIR) spectrum, multispectral imaging across multiple NIR bands provides complementary reflectance information that can enhance the generalizability of PAD methods. In this work, we propose \textbf{SpectraIrisPAD}, a novel deep learning-based framework for robust multispectral iris PAD. The SpectraIrisPAD leverages a DINOv2 Vision Transformer (ViT) backbone equipped with learnable spectral positional encoding, token fusion, and contrastive learning to extract discriminative, band-specific features that effectively distinguish bona fide samples from various spoofing artifacts. Furthermore, we introduce a new comprehensive dataset Multispectral Iris PAD (\textbf{MSIrPAD}) with diverse PAIs, captured using a custom-designed multispectral iris sensor operating at five distinct NIR wavelengths (800\,nm, 830\,nm, 850\,nm, 870\,nm, and 980\,nm). The dataset includes 18,848 iris images encompassing eight diverse PAI categories, including five textured contact lenses, print attacks, and display-based attacks. We conduct comprehensive experiments under unseen attack evaluation protocols to assess the generalization capability of the proposed method. SpectraIrisPAD consistently outperforms several state-of-the-art baselines across all performance metrics, demonstrating superior robustness and generalizability in detecting a wide range of presentation attacks.

---

### JaxWildfire: A GPU-Accelerated Wildfire Simulator for Reinforcement Learning

Authors: Ufuk Çakır, Victor-Alexandru Darvariu, Bruno Lacerda, Nick Hawes
Published: 2025-12-05
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.06102v1

Artificial intelligence methods are increasingly being explored for managing wildfires and other natural hazards. In particular, reinforcement learning (RL) is a promising path towards improving outcomes in such uncertain decision-making scenarios and moving beyond reactive strategies. However, training RL agents requires many environment interactions, and the speed of existing wildfire simulators is a severely limiting factor. We introduce $\texttt{JaxWildfire}$, a simulator underpinned by a principled probabilistic fire spread model based on cellular automata. It is implemented in JAX and enables vectorized simulations using $\texttt{vmap}$, allowing high throughput of simulations on GPUs. We demonstrate that $\texttt{JaxWildfire}$ achieves 6-35x speedup over existing software and enables gradient-based optimization of simulator parameters. Furthermore, we show that $\texttt{JaxWildfire}$ can be used to train RL agents to learn wildfire suppression policies. Our work is an important step towards enabling the advancement of RL techniques for managing natural hazards.

---

### Empathy by Design: Aligning Large Language Models for Healthcare Dialogue

Authors: Emre Umucu, Guillermina Solis, Leon Garza, Emilia Rivas, Beatrice Lee, Anantaa Kotal, Aritran Piplai
Published: 2025-12-05
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2512.06097v1

General-purpose large language models (LLMs) have demonstrated remarkable generative and reasoning capabilities but remain limited in healthcare and caregiving applications due to two key deficiencies: factual unreliability and a lack of empathetic communication. These shortcomings pose significant risks in sensitive contexts where users, particularly non-professionals and caregivers, seek medically relevant guidance or emotional reassurance. To address these challenges, we introduce a Direct Preference Optimization (DPO)-based alignment framework designed to improve factual correctness, semantic coherence, and human-centric qualities such as empathy, politeness, and simplicity in caregiver-patient dialogues. Our approach fine-tunes domain-adapted LLMs using pairwise preference data, where preferred responses reflect supportive and accessible communication styles while rejected ones represent prescriptive or overly technical tones. This direct optimization method aligns model outputs with human preferences more efficiently than traditional reinforcement-learning-based alignment. Empirical evaluations across multiple open and proprietary LLMs show that our DPO-tuned models achieve higher semantic alignment, improved factual accuracy, and stronger human-centric evaluation scores compared to baseline and commercial alternatives such as Google medical dialogue systems. These improvements demonstrate that preference-based alignment offers a scalable and transparent pathway toward developing trustworthy, empathetic, and clinically informed AI assistants for caregiver and healthcare communication. Our open-source code is available at: https://github.com/LeonG19/Empathy-by-Design

---

### BeLLA: End-to-End Birds Eye View Large Language Assistant for Autonomous Driving

Authors: Karthik Mohan, Sonam Singh, Amit Arvind Kale
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.06096v1

The rapid development of Vision-Language models (VLMs) and Multimodal Language Models (MLLMs) in autonomous driving research has significantly reshaped the landscape by enabling richer scene understanding, context-aware reasoning, and more interpretable decision-making. However, a lot of existing work often relies on either single-view encoders that fail to exploit the spatial structure of multi-camera systems or operate on aggregated multi-view features, which lack a unified spatial representation, making it more challenging to reason about ego-centric directions, object relations, and the wider context. We thus present BeLLA, an end-to-end architecture that connects unified 360° BEV representations with a large language model for question answering in autonomous driving. We primarily evaluate our work using two benchmarks - NuScenes-QA and DriveLM, where BeLLA consistently outperforms existing approaches on questions that require greater spatial reasoning, such as those involving relative object positioning and behavioral understanding of nearby objects, achieving up to +9.3% absolute improvement in certain tasks. In other categories, BeLLA performs competitively, demonstrating the capability of handling a diverse range of questions.

---

### Shoot-Bounce-3D: Single-Shot Occlusion-Aware 3D from Lidar by Decomposing Two-Bounce Light

Authors: Tzofi Klinghoffer, Siddharth Somasundaram, Xiaoyu Xiang, Yuchen Fan, Christian Richardt, Akshat Dave, Ramesh Raskar, Rakesh Ranjan
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.06080v1

3D scene reconstruction from a single measurement is challenging, especially in the presence of occluded regions and specular materials, such as mirrors. We address these challenges by leveraging single-photon lidars. These lidars estimate depth from light that is emitted into the scene and reflected directly back to the sensor. However, they can also measure light that bounces multiple times in the scene before reaching the sensor. This multi-bounce light contains additional information that can be used to recover dense depth, occluded geometry, and material properties. Prior work with single-photon lidar, however, has only demonstrated these use cases when a laser sequentially illuminates one scene point at a time. We instead focus on the more practical - and challenging - scenario of illuminating multiple scene points simultaneously. The complexity of light transport due to the combined effects of multiplexed illumination, two-bounce light, shadows, and specular reflections is challenging to invert analytically. Instead, we propose a data-driven method to invert light transport in single-photon lidar. To enable this approach, we create the first large-scale simulated dataset of ~100k lidar transients for indoor scenes. We use this dataset to learn a prior on complex light transport, enabling measured two-bounce light to be decomposed into the constituent contributions from each laser spot. Finally, we experimentally demonstrate how this decomposed light can be used to infer 3D geometry in scenes with occlusions and mirrors from a single measurement. Our code and dataset are released at https://shoot-bounce-3d.github.io.

---

### EgoEdit: Dataset, Real-Time Streaming Model, and Benchmark for Egocentric Video Editing

Authors: Runjia Li, Moayed Haji-Ali, Ashkan Mirzaei, Chaoyang Wang, Arpit Sahni, Ivan Skorokhodov, Aliaksandr Siarohin, Tomas Jakab, Junlin Han, Sergey Tulyakov, Philip Torr, Willi Menapace
Published: 2025-12-05
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2512.06065v1

We study instruction-guided editing of egocentric videos for interactive AR applications. While recent AI video editors perform well on third-person footage, egocentric views present unique challenges - including rapid egomotion and frequent hand-object interactions - that create a significant domain gap. Moreover, existing offline editing pipelines suffer from high latency, limiting real-time interaction. To address these issues, we present a complete ecosystem for egocentric video editing. First, we construct EgoEditData, a carefully designed and manually curated dataset specifically designed for egocentric editing scenarios, featuring rich hand-object interactions, while explicitly preserving hands. Second, we develop EgoEdit, an instruction-following egocentric video editor that supports real-time streaming inference on a single GPU. Finally, we introduce EgoEditBench, an evaluation suite targeting instruction faithfulness, hand and interaction preservation, and temporal stability under egomotion. Across both egocentric and general editing tasks, EgoEdit produces temporally stable, instruction-faithful results with interactive latency. It achieves clear gains on egocentric editing benchmarks-where existing methods struggle-while maintaining performance comparable to the strongest baselines on general editing tasks. EgoEditData and EgoEditBench will be made public for the research community. See our website at https://snap-research.github.io/EgoEdit

---

### When Privacy Isn't Synthetic: Hidden Data Leakage in Generative AI Models

Authors: S. M. Mustaqim, Anantaa Kotal, Paul H. Yi
Published: 2025-12-05
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.06062v1

Generative models are increasingly used to produce privacy-preserving synthetic data as a safe alternative to sharing sensitive training datasets. However, we demonstrate that such synthetic releases can still leak information about the underlying training samples through structural overlap in the data manifold. We propose a black-box membership inference attack that exploits this vulnerability without requiring access to model internals or real data. The attacker repeatedly queries the generative model to obtain large numbers of synthetic samples, performs unsupervised clustering to identify dense regions of the synthetic distribution, and then analyzes cluster medoids and neighborhoods that correspond to high-density regions in the original training data. These neighborhoods act as proxies for training samples, enabling the adversary to infer membership or reconstruct approximate records. Our experiments across healthcare, finance, and other sensitive domains show that cluster overlap between real and synthetic data leads to measurable membership leakage-even when the generator is trained with differential privacy or other noise mechanisms. The results highlight an under-explored attack surface in synthetic data generation pipelines and call for stronger privacy guarantees that account for distributional neighborhood inference rather than sample-level memorization alone, underscoring its role in privacy-preserving data publishing. Implementation and evaluation code are publicly available at:github.com/Cluster-Medoid-Leakage-Attack.

---

### Reinforcement Learning Integrated Agentic RAG for Software Test Cases Authoring

Authors: Mohanakrishnan Hariharan
Published: 2025-12-05
Categories: cs.SE, cs.AI
arXiv: https://arxiv.org/abs/2512.06060v1

This paper introduces a framework that integrates reinforcement learning (RL) with autonomous agents to enable continuous improvement in the automated process of software test cases authoring from business requirement documents within Quality Engineering (QE) workflows. Conventional systems employing Large Language Models (LLMs) generate test cases from static knowledge bases, which fundamentally limits their capacity to enhance performance over time. Our proposed Reinforcement Infused Agentic RAG (Retrieve, Augment, Generate) framework overcomes this limitation by employing AI agents that learn from QE feedback, assessments, and defect discovery outcomes to automatically improve their test case generation strategies. The system combines specialized agents with a hybrid vector-graph knowledge base that stores and retrieves software testing knowledge. Through advanced RL algorithms, specifically Proximal Policy Optimization (PPO) and Deep Q-Networks (DQN), these agents optimize their behavior based on QE-reported test effectiveness, defect detection rates, and workflow metrics. As QEs execute AI-generated test cases and provide feedback, the system learns from this expert guidance to improve future iterations. Experimental validation on enterprise Apple projects yielded substantive improvements: a 2.4% increase in test generation accuracy (from 94.8% to 97.2%), and a 10.8% improvement in defect detection rates. The framework establishes a continuous knowledge refinement loop driven by QE expertise, resulting in progressively superior test case quality that enhances, rather than replaces, human testing capabilities.

---

### Deep learning recognition and analysis of Volatile Organic Compounds based on experimental and synthetic infrared absorption spectra

Authors: Andrea Della Valle, Annalisa D'Arco, Tiziana Mancini, Rosanna Mosetti, Maria Chiara Paolozzi, Stefano Lupi, Sebastiano Pilati, Andrea Perali
Published: 2025-12-05
Categories: cs.LG, physics.app-ph, physics.chem-ph
arXiv: https://arxiv.org/abs/2512.06059v1

Volatile Organic Compounds (VOCs) are organic molecules that have low boiling points and therefore easily evaporate into the air. They pose significant risks to human health, making their accurate detection the crux of efforts to monitor and minimize exposure. Infrared (IR) spectroscopy enables the ultrasensitive detection at low-concentrations of VOCs in the atmosphere by measuring their IR absorption spectra. However, the complexity of the IR spectra limits the possibility to implement VOC recognition and quantification in real-time. While deep neural networks (NNs) are increasingly used for the recognition of complex data structures, they typically require massive datasets for the training phase. Here, we create an experimental VOC dataset for nine different classes of compounds at various concentrations, using their IR absorption spectra. To further increase the amount of spectra and their diversity in term of VOC concentration, we augment the experimental dataset with synthetic spectra created via conditional generative NNs. This allows us to train robust discriminative NNs, able to reliably identify the nine VOCs, as well as to precisely predict their concentrations. The trained NN is suitable to be incorporated into sensing devices for VOCs recognition and analysis.

---

### Representation Learning for Point Cloud Understanding

Authors: Siming Yan
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.06058v1

With the rapid advancement of technology, 3D data acquisition and utilization have become increasingly prevalent across various fields, including computer vision, robotics, and geospatial analysis. 3D data, captured through methods such as 3D scanners, LiDARs, and RGB-D cameras, provides rich geometric, shape, and scale information. When combined with 2D images, 3D data offers machines a comprehensive understanding of their environment, benefiting applications like autonomous driving, robotics, remote sensing, and medical treatment. This dissertation focuses on three main areas: supervised representation learning for point cloud primitive segmentation, self-supervised learning methods, and transfer learning from 2D to 3D. Our approach, which integrates pre-trained 2D models to support 3D network training, significantly improves 3D understanding without merely transforming 2D data. Extensive experiments validate the effectiveness of our methods, showcasing their potential to advance point cloud representation learning by effectively integrating 2D knowledge.

---

### The Road of Adaptive AI for Precision in Cybersecurity

Authors: Sahil Garg
Published: 2025-12-05
Categories: cs.CR, cs.AI, cs.CL, cs.LG, cs.SE
arXiv: https://arxiv.org/abs/2512.06048v1

Cybersecurity's evolving complexity presents unique challenges and opportunities for AI research and practice. This paper shares key lessons and insights from designing, building, and operating production-grade GenAI pipelines in cybersecurity, with a focus on the continual adaptation required to keep pace with ever-shifting knowledge bases, tooling, and threats. Our goal is to provide an actionable perspective for AI practitioners and industry stakeholders navigating the frontier of GenAI for cybersecurity, with particular attention to how different adaptation mechanisms complement each other in end-to-end systems. We present practical guidance derived from real-world deployments, propose best practices for leveraging retrieval- and model-level adaptation, and highlight open research directions for making GenAI more robust, precise, and auditable in cyber defense.

---

### Beyond Prototyping: Autonomous, Enterprise-Grade Frontend Development from Pixel to Production via a Specialized Multi-Agent Framework

Authors: Ramprasath Ganesaraja, Swathika N, Saravanan AP, Kamalkumar Rathinasamy, Chetana Amancharla, Rahul Das, Sahil Dilip Panse, Aditya Batwe, Dileep Vijayan, Veena Ashok, Thanushree A P, Kausthubh J Rao et al.
Published: 2025-12-05
Categories: cs.SE, cs.AI
arXiv: https://arxiv.org/abs/2512.06046v1

We present AI4UI, a framework of autonomous front-end development agents purpose-built to meet the rigorous requirements of enterprise-grade application delivery. Unlike general-purpose code assistants designed for rapid prototyping, AI4UI focuses on production readiness delivering secure, scalable, compliant, and maintainable UI code integrated seamlessly into enterprise workflows. AI4UI operates with targeted human-in-the-loop involvement: at the design stage, developers embed a Gen-AI-friendly grammar into Figma prototypes to encode requirements for precise interpretation; and at the post processing stage, domain experts refine outputs for nuanced design adjustments, domain-specific optimizations, and compliance needs. Between these stages, AI4UI runs fully autonomously, converting designs into engineering-ready UI code. Technical contributions include a Figma grammar for autonomous interpretation, domain-aware knowledge graphs, a secure abstract/package code integration strategy, expertise driven architecture templates, and a change-oriented workflow coordinated by specialized agent roles. In large-scale benchmarks against industry baselines and leading competitor systems, AI4UI achieved 97.24% platform compatibility, 87.10% compilation success, 86.98% security compliance, 78.00% feature implementation success, 73.50% code-review quality, and 73.36% UI/UX consistency. In blind preference studies with 200 expert evaluators, AI4UI emerged as one of the leaders demonstrating strong competitive standing among leading solutions. Operating asynchronously, AI4UI generates thousands of validated UI screens in weeks rather than months, compressing delivery timeline

---

### Auto-SPT: Automating Semantic Preserving Transformations for Code

Authors: Ashish Hooda, Mihai Christodorescu, Chuangang Ren, Aaron Wilson, Kassem Fawaz, Somesh Jha
Published: 2025-12-05
Categories: cs.SE, cs.AI
arXiv: https://arxiv.org/abs/2512.06042v1

Machine learning (ML) models for code clone detection determine whether two pieces of code are semantically equivalent, which in turn is a key building block for software-engineering tasks like refactoring and security tasks like vulnerability and malware detection. While these models are predominantly trained on clean, structured code datasets, real-world code often undergoes a variety of semantic-preserving transformations, including refactoring, minification, automated formatting, and compiler optimizations. To address this critical gap between training and test data, we propose Auto-SPT, a novel framework to automatically construct synthetic-data generators for code. Auto-SPT is designed to produce Semantic Preserving Transformations (SPTs) that alter a program's syntactic structure while preserving its functionality and is instantiated on top of Large Language Models (LLMs). In particular, we use LLMs to craft a diverse set of SPTs, generate strong implementations for these SPTs, and compose them to result into strong transformations. Our formal analysis shows that the diversity of SPTs impacts the strength of their composition. We then empirically demonstrate that Auto-SPT generates more diverse SPTs than existing approaches and these SPTs significantly drop the performance of state-of-the-art code clone detectors. Further experiments show Auto-SPT can be used to enhance code datasets for training, to produce code-clone detection models that are robust to real-world, adversarial code transformations.

---

### Enhancing Retrieval-Augmented Generation with Entity Linking for Educational Platforms

Authors: Francesco Granata, Francesco Poggi, Misael Mongiovì
Published: 2025-12-05
Categories: cs.IR, cs.AI, cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2512.05967v2

In the era of Large Language Models (LLMs), Retrieval-Augmented Generation (RAG) architectures are gaining significant attention for their ability to ground language generation in reliable knowledge sources. Despite their effectiveness, RAG systems based solely on semantic similarity often fail to ensure factual accuracy in specialized domains, where terminological ambiguity can affect retrieval relevance. This study proposes ELERAG, an enhanced RAG architecture that integrates a factual signal derived from Entity Linking to improve the accuracy of educational question-answering systems in Italian. The system includes a Wikidata-based Entity Linking module and implements a hybrid re-ranking strategy based on Reciprocal Rank Fusion (RRF). To validate our approach, we compared it against standard baselines and state-of-the-art methods, including a Weighted-Score Re-ranking, a standalone Cross-Encoder and a combined RRF+Cross-Encoder pipeline. Experiments were conducted on two benchmarks: a custom academic dataset and the standard SQuAD-it dataset. Results show that, in domain-specific contexts, ELERAG significantly outperforms both the baseline and the Cross-Encoder configurations. Conversely, the Cross-Encoder approaches achieve the best results on the general-domain dataset. These findings provide strong experimental evidence of the domain mismatch effect, highlighting the importance of domain-adapted hybrid strategies to enhance factual precision in educational RAG systems without relying on computationally expensive models trained on disparate data distributions. They also demonstrate the potential of entity-aware RAG systems in educational environments, fostering adaptive and reliable AI-based tutoring tools.

---

### Enhancing Retrieval-Augmented Generation with Entity Linking for Educational Platforms

Authors: Francesco Granata, Francesco Poggi, Misael Mongiovì
Published: 2025-12-05
Categories: cs.IR, cs.AI, cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2512.05967v1

In the era of Large Language Models (LLMs), Retrieval-Augmented Generation (RAG) architectures are gaining significant attention for their ability to ground language generation in reliable knowledge sources. Despite their impressive effectiveness in many areas, RAG systems based solely on semantic similarity often fail to ensure factual accuracy in specialized domains, where terminological ambiguity can affect retrieval relevance. This study proposes an enhanced RAG architecture that integrates a factual signal derived from Entity Linking to improve the accuracy of educational question-answering systems in Italian. The system includes a Wikidata-based Entity Linking module and implements three re-ranking strategies to combine semantic and entity-based information: a hybrid score weighting model, reciprocal rank fusion, and a cross-encoder re-ranker. Experiments were conducted on two benchmarks: a custom academic dataset and the standard SQuAD-it dataset. Results show that, in domain-specific contexts, the hybrid schema based on reciprocal rank fusion significantly outperforms both the baseline and the cross-encoder approach, while the cross-encoder achieves the best results on the general-domain dataset. These findings confirm the presence of an effect of domain mismatch and highlight the importance of domain adaptation and hybrid ranking strategies to enhance factual precision and reliability in retrieval-augmented generation. They also demonstrate the potential of entity-aware RAG systems in educational environments, fostering adaptive and reliable AI-based tutoring tools.

---

### EditThinker: Unlocking Iterative Reasoning for Any Image Editor

Authors: Hongyu Li, Manyuan Zhang, Dian Zheng, Ziyu Guo, Yimeng Jia, Kaituo Feng, Hao Yu, Yexin Liu, Yan Feng, Peng Pei, Xunliang Cai, Linjiang Huang et al.
Published: 2025-12-05
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.05965v1

Instruction-based image editing has emerged as a prominent research area, which, benefiting from image generation foundation models, have achieved high aesthetic quality, making instruction-following capability the primary challenge. Existing approaches improve instruction adherence via supervised or reinforcement learning, yet single-turn success rates remain limited due to inherent stochasticity and a lack of deliberation. In this work, we propose a deliberative editing framework to 'think' while they edit, which simulates the human cognitive loop by iteratively executing a Think-while-Edit cycle: Critiquing results and Refining instructions , followed by Repeating the generation until satisfactory. Specifically, we train a single MLLM, EditThinker, to act as the reasoning engine of this framework, which jointly produce the critique score, reasoning process, and refined instructions. We employ reinforcement learning to align the EditThinker's thinking with its editing, thereby generating more targeted instruction improvements. Extensive experiments on four benchmarks demonstrate that our approach significantly improves the instruction-following capability of any image editing model by a large margin. We will release our data construction framework, datasets, and models to benefit the community.

---

### Training-Time Action Conditioning for Efficient Real-Time Chunking

Authors: Kevin Black, Allen Z. Ren, Michael Equi, Sergey Levine
Published: 2025-12-05
Categories: cs.RO, cs.AI
arXiv: https://arxiv.org/abs/2512.05964v2

Real-time chunking (RTC) enables vision-language-action models (VLAs) to generate smooth, reactive robot trajectories by asynchronously predicting action chunks and conditioning on previously committed actions via inference-time inpainting. However, this inpainting method introduces computational overhead that increases inference latency. In this work, we propose a simple alternative: simulating inference delay at training time and conditioning on action prefixes directly, eliminating any inference-time overhead. Our method requires no modifications to the model architecture or robot runtime, and can be implemented with only a few additional lines of code. In simulated experiments, we find that training-time RTC outperforms inference-time RTC at higher inference delays. In real-world experiments on box building and espresso making tasks with the $π_{0.6}$ VLA, we demonstrate that training-time RTC maintains both task performance and speed parity with inference-time RTC while being computationally cheaper. Our results suggest that training-time action conditioning is a practical drop-in replacement for inference-time inpainting in real-time robot control.

---

### Training-Time Action Conditioning for Efficient Real-Time Chunking

Authors: Kevin Black, Allen Z. Ren, Michael Equi, Sergey Levine
Published: 2025-12-05
Categories: cs.RO, cs.AI
arXiv: https://arxiv.org/abs/2512.05964v1

Real-time chunking (RTC) enables vision-language-action models (VLAs) to generate smooth, reactive robot trajectories by asynchronously predicting action chunks and conditioning on previously committed actions via inference-time inpainting. However, this inpainting method introduces computational overhead that increases inference latency. In this work, we propose a simple alternative: simulating inference delay at training time and conditioning on action prefixes directly, eliminating any inference-time overhead. Our method requires no modifications to the model architecture or robot runtime, and can be implemented with only a few additional lines of code. In simulated experiments, we find that training-time RTC outperforms inference-time RTC at higher inference delays. In real-world experiments on box building and espresso making tasks with the $π_{0.6}$ VLA, we demonstrate that training-time RTC maintains both task performance and speed parity with inference-time RTC while being computationally cheaper. Our results suggest that training-time action conditioning is a practical drop-in replacement for inference-time inpainting in real-time robot control.

---

### Whatever Remains Must Be True: Filtering Drives Reasoning in LLMs, Shaping Diversity

Authors: Germán Kruszewski, Pierre Erbacher, Jos Rozen, Marc Dymetman
Published: 2025-12-05
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.05962v2

Reinforcement Learning (RL) has become the de facto standard for tuning LLMs to solve tasks involving reasoning. However, growing evidence shows that models trained in such way often suffer from a significant loss in diversity. We argue that this arises because RL implicitly optimizes the "mode-seeking" or "zero-forcing" Reverse KL to a target distribution causing the model to concentrate mass on certain high-probability regions of the target while neglecting others. In this work, we instead begin from an explicit target distribution, obtained by filtering out incorrect answers while preserving the relative probabilities of correct ones. Starting from a pre-trained LLM, we approximate this target distribution using the $α$-divergence family, which unifies prior approaches and enables direct control of the precision-diversity trade-off by interpolating between mode-seeking and mass-covering divergences. On a Lean theorem-proving benchmark, our method achieves state-of-the-art performance along the coverage-precision Pareto frontier, outperforming all prior methods on the coverage axis.

---

### Whatever Remains Must Be True: Filtering Drives Reasoning in LLMs, Shaping Diversity

Authors: Germán Kruszewski, Pierre Erbacher, Jos Rozen, Marc Dymetman
Published: 2025-12-05
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.05962v1

Reinforcement Learning (RL) has become the de facto standard for tuning LLMs to solve tasks involving reasoning. However, growing evidence shows that models trained in such way often suffer from a significant loss in diversity. We argue that this arises because RL implicitly optimizes the "mode-seeking" or "zero-forcing" Reverse KL to a target distribution causing the model to concentrate mass on certain high-probability regions of the target while neglecting others. In this work, we instead begin from an explicit target distribution, obtained by filtering out incorrect answers while preserving the relative probabilities of correct ones. Starting from a pre-trained LLM, we approximate this target distribution using the $α$-divergence family, which unifies prior approaches and enables direct control of the precision-diversity trade-off by interpolating between mode-seeking and mass-covering divergences. On a Lean theorem-proving benchmark, our method achieves state-of-the-art performance along the coverage-precision Pareto frontier, outperforming all prior methods on the coverage axis.

---

### AQUA-Net: Adaptive Frequency Fusion and Illumination Aware Network for Underwater Image Enhancement

Authors: Munsif Ali, Najmul Hassan, Lucia Ventura, Davide Di Bari, Simonepietro Canese
Published: 2025-12-05
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2512.05960v1

Underwater images often suffer from severe color distortion, low contrast, and a hazy appearance due to wavelength-dependent light absorption and scattering. Simultaneously, existing deep learning models exhibit high computational complexity, which limits their practical deployment for real-time underwater applications. To address these challenges, this paper presents a novel underwater image enhancement model, called Adaptive Frequency Fusion and Illumination Aware Network (AQUA-Net). It integrates a residual encoder decoder with dual auxiliary branches, which operate in the frequency and illumination domains. The frequency fusion encoder enriches spatial representations with frequency cues from the Fourier domain and preserves fine textures and structural details. Inspired by Retinex, the illumination-aware decoder performs adaptive exposure correction through a learned illumination map that separates reflectance from lighting effects. This joint spatial, frequency, and illumination design enables the model to restore color balance, visual contrast, and perceptual realism under diverse underwater conditions. Additionally, we present a high-resolution, real-world underwater video-derived dataset from the Mediterranean Sea, which captures challenging deep-sea conditions with realistic visual degradations to enable robust evaluation and development of deep learning models. Extensive experiments on multiple benchmark datasets show that AQUA-Net performs on par with SOTA in both qualitative and quantitative evaluations while using less number of parameters. Ablation studies further confirm that the frequency and illumination branches provide complementary contributions that improve visibility and color representation. Overall, the proposed model shows strong generalization capability and robustness, and it provides an effective solution for real-world underwater imaging applications.

---

### M4-RAG: A Massive-Scale Multilingual Multi-Cultural Multimodal RAG

Authors: David Anugraha, Patrick Amadeus Irawan, Anshul Singh, En-Shiun Annie Lee, Genta Indra Winata
Published: 2025-12-05
Categories: cs.CL, cs.AI, cs.CV
arXiv: https://arxiv.org/abs/2512.05959v2

Vision-language models (VLMs) have achieved strong performance in visual question answering (VQA), yet they remain constrained by static training data. Retrieval-Augmented Generation (RAG) mitigates this limitation by enabling access to up-to-date, culturally grounded, and multilingual information; however, multilingual multimodal RAG remains largely underexplored. We introduce M4-RAG, a massive-scale benchmark spanning 42 languages, 56 regional dialects and registers, and 189 countries, comprising over 80,000 culturally diverse image-question pairs for evaluating retrieval-augmented VQA across languages and modalities. To balance realism with reproducibility, we build a controlled retrieval environment containing millions of carefully curated multilingual documents relevant to the query domains, approximating real-world retrieval conditions while ensuring consistent experimentation. Our systematic evaluation reveals that although RAG consistently benefits smaller VLMs, it fails to scale to larger models and often even degrades their performance, exposing a critical mismatch between model size and current retrieval effectiveness. Our cross-lingual evaluations also reveal significant performance degradation when prompts or retrieved context are provided in non-English languages. The code, datasets, and evaluation protocols for M4-RAG are available as open-source at https://github.com/davidanugraha/M4-RAG.

---

### M4-RAG: A Massive-Scale Multilingual Multi-Cultural Multimodal RAG

Authors: David Anugraha, Patrick Amadeus Irawan, Anshul Singh, En-Shiun Annie Lee, Genta Indra Winata
Published: 2025-12-05
Categories: cs.CL, cs.AI, cs.CV
arXiv: https://arxiv.org/abs/2512.05959v1

Vision-language models (VLMs) have achieved strong performance in visual question answering (VQA), yet they remain constrained by static training data. Retrieval-Augmented Generation (RAG) mitigates this limitation by enabling access to up-to-date, culturally grounded, and multilingual information; however, multilingual multimodal RAG remains largely underexplored. We introduce M4-RAG, a massive-scale benchmark covering 42 languages and 56 regional dialects and registers, comprising over 80,000 culturally diverse image-question pairs for evaluating retrieval-augmented VQA across languages and modalities. To balance realism with reproducibility, we build a controlled retrieval environment containing millions of carefully curated multilingual documents relevant to the query domains, approximating real-world retrieval conditions while ensuring consistent experimentation. Our systematic evaluation reveals that although RAG consistently benefits smaller VLMs, it fails to scale to larger models and often even degrades their performance, exposing a critical mismatch between model size and current retrieval effectiveness. M4-RAG provides a foundation for advancing next-generation RAG systems capable of reasoning seamlessly across languages, modalities, and cultural contexts.

---

### MaxShapley: Towards Incentive-compatible Generative Search with Fair Context Attribution

Authors: Sara Patel, Mingxun Zhou, Giulia Fanti
Published: 2025-12-05
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.05958v2

Generative search engines based on large language models (LLMs) are replacing traditional search, fundamentally changing how information providers are compensated. To sustain this ecosystem, we need fair mechanisms to attribute and compensate content providers based on their contributions to generated answers. We introduce MaxShapley, an efficient algorithm for fair credit attribution in generative search pipelines that retrieve external sources before generation. MaxShapley is a special case of the celebrated Shapley value; it leverages a de-composable max-sum utility function to compute attributions with polynomial-time computation in the number of documents, as opposed to the exponential cost of Shapley values. We evaluate MaxShapley on three multi-hop QA datasets (HotPotQA, MuSiQUE, MS MARCO); MaxShapley achieves comparable attribution quality to exact Shapley computation, while consuming a fraction of its tokens--for instance, it gives up to a 9x reduction in resource consumption over prior state-of-the-art methods at the same attribution accuracy. We release open-source code and re-calibrated datasets. An educational demo is available at https://fair-search.com.

---

### MaxShapley: Towards Incentive-compatible Generative Search with Fair Context Attribution

Authors: Sara Patel, Mingxun Zhou, Giulia Fanti
Published: 2025-12-05
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.05958v1

Generative search engines based on large language models (LLMs) are replacing traditional search, fundamentally changing how information providers are compensated. To sustain this ecosystem, we need fair mechanisms to attribute and compensate content providers based on their contributions to generated answers. We introduce MaxShapley, an efficient algorithm for fair attribution in generative search pipelines that use retrieval-augmented generation (RAG). MaxShapley is a special case of the celebrated Shapley value; it leverages a decomposable max-sum utility function to compute attributions with linear computation in the number of documents, as opposed to the exponential cost of Shapley values. We evaluate MaxShapley on three multi-hop QA datasets (HotPotQA, MuSiQUE, MS MARCO); MaxShapley achieves comparable attribution quality to exact Shapley computation, while consuming a fraction of its tokens--for instance, it gives up to an 8x reduction in resource consumption over prior state-of-the-art methods at the same attribution accuracy.

---

### Consequences of Kernel Regularity for Bandit Optimization

Authors: Madison Lee, Tara Javidi
Published: 2025-12-05
Categories: stat.ML, cs.LG
arXiv: https://arxiv.org/abs/2512.05957v1

In this work we investigate the relationship between kernel regularity and algorithmic performance in the bandit optimization of RKHS functions. While reproducing kernel Hilbert space (RKHS) methods traditionally rely on global kernel regressors, it is also common to use a smoothness-based approach that exploits local approximations. We show that these perspectives are deeply connected through the spectral properties of isotropic kernels. In particular, we characterize the Fourier spectra of the Matérn, square-exponential, rational-quadratic, $γ$-exponential, piecewise-polynomial, and Dirichlet kernels, and show that the decay rate determines asymptotic regret from both viewpoints. For kernelized bandit algorithms, spectral decay yields upper bounds on the maximum information gain, governing worst-case regret, while for smoothness-based methods, the same decay rates establish Hölder space embeddings and Besov space norm-equivalences, enabling local continuity analysis. These connections show that kernel-based and locally adaptive algorithms can be analyzed within a unified framework. This allows us to derive explicit regret bounds for each kernel family, obtaining novel results in several cases and providing improved analysis for others. Furthermore, we analyze LP-GP-UCB, an algorithm that combines both approaches, augmenting global Gaussian process surrogates with local polynomial estimators. While the hybrid approach does not uniformly dominate specialized methods, it achieves order-optimality across multiple kernel families.

---