# Research Papers — 2025-12 · bundle page 123 of 175 > Full-text markdown of papers #6101–6150 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/123 [← Previous bundle](https://aipod.dev/papers/archive/2025-12/page/122.md) [Next bundle →](https://aipod.dev/papers/archive/2025-12/page/124.md) ### HistCAD: Constraint-Aware Parametric CAD Histories for Evaluating Editability Authors: Xintong Dong, Chuanyang Li, Peng Zheng, Chuqi Han, Jiaxin Jing, Hailong Shen, Yanzhi Song, Zhouwang Yang Published: 2025-12-08 Categories: cs.GR, cs.AI arXiv: https://arxiv.org/abs/2602.19171v4 Sketch constraints specify geometric conditions for constructing and modifying parametric CAD models. We study whether predicted constraints allow a given history to reproduce the required initial model and support prescribed dimensional edits. We introduce HistCAD, an executable representation and dataset whose Academic and Industrial collections contain 180,495 parametric construction histories with entity-referenced sketch constraints and retained feature operations. Predictors receive these histories with the geometry and feature definitions retained and explicit sketch constraints removed. They generate constraints for every sketch without seeing the edit request. The benchmark compares models built with alternative constraint sets for the same history under the same dimensional edit. An edit succeeds when the model reproduces the required initial geometry, reaches the target value, preserves specified relations and unedited dimensions in the target sketch, and rebuilds through the complete history. A predictor trained on both collections and supplied with descriptions of the input histories achieves overall edit success of 52.4% on Academic and 29.0% on Industrial. Models retaining only endpoint-connectivity constraints in the target sketch and the history's constraints elsewhere can reach the target and rebuild while failing preservation. For all-sketch predictions, we retain the target-sketch prediction and restore the history's constraints in other sketches. More models then reproduce the required initial geometry, and some of these newly matched models complete the edit. HistCAD connects constraint learning to the construction and revision of parametric CAD models. --- ### Response-Based Knowledge Distillation for Multilingual Jailbreak Prevention Unwittingly Compromises Safety Authors: Max Zhang, Derek Liu, Kai Zhang, Joshua Franco, Haihao Liu Published: 2025-12-08 Categories: cs.CL arXiv: https://arxiv.org/abs/2602.11157v2 Large language models (LLMs) are increasingly deployed worldwide, yet their safety alignment remains predominantly English-centric. This allows for vulnerabilities in non-English contexts, especially with low-resource languages. We introduce a novel application of knowledge distillation (KD) in the context of multilingual jailbreak prevention, examining its efficacy. We distill the refusal behaviors of a proprietary teacher model (OpenAI o1-mini) with Low-Rank Adaptation (LoRA) into three open-source student models: Meta-Llama-3-8B-Instruct, Gemma-2-2B-IT, and Qwen3-8B, using ~28,000 multilingual jailbreak prompts from XSafety via black-box response-based, parameter-efficient fine-tuning (PEFT). Evaluation on the MultiJail benchmark reveals a counterintuitive behavior: standard fine-tuning on the teacher's ``safe'' refusal data inadvertently increases Jailbreak Success Rate (JSR) for all student models, up to 16.6 percentage points. Our experiments reveal a divergent generalization to unseen languages during distillation, with varying outcomes depending on the base model. By removing a primary source of safety degradation, nuanced `boundary' refusals, we mitigate or even reverse safety declines in student models, although reductions in reasoning performance (GSM8K) persist. Overall, our exploratory study highlights the challenges and potential of KD as a technique for multilingual safety alignment, offering a foundation for future research in this direction. --- ### GTAC: A Generative Transformer for Approximate Circuits Authors: Jingxin Wang, Shitong Guo, Wenhui Liang, Ruicheng Dai, Ruogu Ding, Xin Ning, Weikang Qian Published: 2025-12-08 Categories: cs.AR, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2601.19906v2 Targeting error-tolerant applications, approximate computing relaxes rigid functional equivalence to significantly improve power, performance, and area. Traditional approximate logic synthesis (ALS) relies on incremental rewriting, limiting design space exploration. Meanwhile, the inherently probabilistic nature of Transformer-based generative AI makes it a natural fit for generating approximate circuits. Exploiting this, we propose GTAC, an end-to-end framework for arbitrary-scale generative ALS. To overcome the memory bottleneck of generative AI, GTAC partitions a large circuit into tractable subcircuits, applies a generative core to produce approximate candidates for each subcircuit, and finally selects proper candidates to form the final design. Its core generative Transformer utilizes a novel irredundant encoding to compactly encode a circuit, alongside a masking mechanism to exclude designs violating the given error bound. Empowered by a self-evolutionary training strategy, GTAC establishes a new paradigm that demonstrates superior performance: It reduces delay by 30.9% and gate count by 50.5% over exact generative baselines and saves 6.5% area with a 4.3x speedup against traditional ALS methods. Furthermore, its irredundant encoding achieves a 33.3x reduction in sequence length and a 61.6x reduction in peak memory compared to conventional memoryless traversal. --- ### Hardware-Aware Model Design and Training of Silicon-based Analog Neural Networks Authors: Giulio Filippeschi, Mirko Brazzini, Cristhopher Mosquera, Marco Lanuzza, Alessandro Catania, Sebastiano Strangio, Giuseppe Iannaccone Published: 2025-12-08 Categories: cs.AR, cs.NE arXiv: https://arxiv.org/abs/2601.19905v1 Silicon-based analog neural networks physically embody the ideal neural network model in an approximate way. We show that by retraining the neural network using a physics-informed hardware-aware model one can fully recover the inference accuracy of the ideal network model even in the presence of significant non-idealities. This is way more promising for scalability and integration density than the default option of improving the fidelity of the analog neural network at the cost of significant energy, area, and design overhead, through extensive calibration and conservative analog design. We first present a physics-informed hardware-aware model for a time-domain vector-matrix multiplier implemented with single-transistor floating-gate memory cells that explicitly accounts for two dominant non-idealities of the physical implementation - capacitive crosstalk and bit-line voltage drop - and integrates seamlessly with modern deep-learning workflows. The model discretizes each operation into adaptive time slots, processes activation patterns in parallel, and accumulates their contributions to predict effective multiplier outputs. Using measurements from a 16x16 silicon array, we calibrate the model, show that crosstalk is layout-dependent and often dominant, and introduce an improved weight-extraction procedure that doubles signal-to-error ratio versus an ideal vector-matrix multiplier model. Finally, we show that by training silicon-based analog neural networks using an hardware-aware model in the forward pass we can recover the accuracy of the ideal software networks across three architectures -- custom MLP on low-resolution MNIST, LeNet-5 on MNIST, and a VGG-style CNN on CIFAR-10 - establishing a complete design-to-deployment workflow for time-domain analog neuromorphic chips. --- ### Ideology as a Problem: Lightweight Logit Steering for Annotator-Specific Alignment in Social Media Analysis Authors: Wei Xia, Haowen Tang, Luozheng Li Published: 2025-12-08 Categories: cs.CL, cs.AI, cs.SI arXiv: https://arxiv.org/abs/2601.04207v1 LLMs internally organize political ideology along low-dimensional structures that are partially, but not fully aligned with human ideological space. This misalignment is systematic, model specific, and measurable. We introduce a lightweight linear probe that both quantifies the misalignment and minimally corrects the output layer. This paper introduces a simple and efficient method for aligning models with specific user opinions. Instead of retraining the model, we calculated a bias score from its internal features and directly adjusted the final output probabilities. This solution is practical and low-cost and preserves the original reasoning power of the model. --- ### LASER: Language Model Regression for Semi-Structured Workflow Resource and Runtime Estimation Authors: Yuxuan Yin, Shengke Zhou, Yunjie Zhang, Ajay Mohindra, Boxun Xu, Peng Li Published: 2025-12-08 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2512.19701v2 Accurate prediction of resource consumption and runtime for cloud workflow jobs is critical for scheduling efficiency, yet remains challenging due to the semi-structured nature of job configurations -- comprising shell commands, tool-specific parameters, dependency graphs, and hierarchical metadata. Traditional ML approaches require brittle feature engineering to flatten this rich information into fixed-size vectors, losing critical semantic context. We present LASER, a framework that fine-tunes LLMs on serialized workflow job configurations for multi-target resource and runtime regression. To address the challenges of numerical regression via generation, we introduce scientific notation output encoding for targets spanning multiple orders of magnitude, and constrained decoding with prefix filling to enforce output validity while reducing inference latency by over 30%. We further show that full-attention fine-tuning improves accuracy over sliding-window LLMs on long job contexts. Validated on large-scale chip design workloads, and GHARuntime, a new public benchmark derived from 580,000+ GitHub Actions runs across 27,000+ repositories, LASER outperforms human experts and SOTA tabular ML baselines, with clear model- and data-scaling behavior, establishing a new paradigm for LLM-based regression on semi-structured workflow data. --- ### A curated UK rain radar data set for training and benchmarking nowcasting models Authors: Viv Atureta, Rifki Priansyah Jasin, Stefan Siegert Published: 2025-12-08 Categories: physics.ao-ph, cs.CV, cs.LG, stat.AP arXiv: https://arxiv.org/abs/2512.17924v1 This paper documents a data set of UK rain radar image sequences for use in statistical modeling and machine learning methods for nowcasting. The main dataset contains 1,000 randomly sampled sequences of length 20 steps (15-minute increments) of 2D radar intensity fields of dimension 40x40 (at 5km spatial resolution). Spatially stratified sampling ensures spatial homogeneity despite removal of clear-sky cases by threshold-based truncation. For each radar sequence, additional atmospheric and geographic features are made available, including date, location, mean elevation, mean wind direction and speed and prevailing storm type. New R functions to extract data from the binary "Nimrod" radar data format are provided. A case study is presented to train and evaluate a simple convolutional neural network for radar nowcasting, including self-contained R code. --- ### Inferring Latent Market Forces: Evaluating LLM Detection of Gamma Exposure Patterns via Obfuscation Testing Authors: Christopher Regan, Ying Xie Published: 2025-12-08 Categories: q-fin.ST, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2512.17923v2 We introduce obfuscation testing, a novel methodology for validating whether large language models detect structural market patterns through causal reasoning rather than temporal association. Testing three dealer hedging constraint patterns (gamma positioning, stock pinning, 0DTE hedging) on 242 trading days (95.6% coverage) of S&P 500 options data, we find LLMs achieve 71.5% detection rate using unbiased prompts that provide only raw gamma exposure values without regime labels or temporal context. The WHO-WHOM-WHAT causal framework forces models to identify the economic actors (dealers), affected parties (directional traders), and structural mechanisms (forced hedging) underlying observed market dynamics. Critically, detection accuracy (91.2%) remains stable even as economic profitability varies quarterly, demonstrating that models identify structural constraints rather than profitable patterns. When prompted with regime labels, detection increases to 100%, but the 71.5% unbiased rate validates genuine pattern recognition. Our findings suggest LLMs possess emergent capabilities for detecting complex financial mechanisms through pure structural reasoning, with implications for systematic strategy development, risk management, and our understanding of how transformer architectures process financial market dynamics. --- ### A Bayesian latent class reinforcement learning framework to capture adaptive, feedback-driven travel behaviour Authors: Georges Sfeir, Stephane Hess, Thomas O. Hancock, Filipe Rodrigues, Jamal Amani Rad, Michiel Bliemer, Matthew Beck, Fayyaz Khan Published: 2025-12-08 Categories: cs.LG, econ.EM, stat.ML arXiv: https://arxiv.org/abs/2512.14713v1 Many travel decisions involve a degree of experience formation, where individuals learn their preferences over time. At the same time, there is extensive scope for heterogeneity across individual travellers, both in their underlying preferences and in how these evolve. The present paper puts forward a Latent Class Reinforcement Learning (LCRL) model that allows analysts to capture both of these phenomena. We apply the model to a driving simulator dataset and estimate the parameters through Variational Bayes. We identify three distinct classes of individuals that differ markedly in how they adapt their preferences: the first displays context-dependent preferences with context-specific exploitative tendencies; the second follows a persistent exploitative strategy regardless of context; and the third engages in an exploratory strategy combined with context-specific preferences. --- ### SepsisSuite: Beyond Risk Stratification -- A Comparative Analysis of Deep Fusion vs. Expert Stacking for Prescriptive Sepsis AI Authors: Ryan Cartularo Published: 2025-12-08 Categories: cs.LG, cs.AI, cs.CL, cs.CV, cs.CY arXiv: https://arxiv.org/abs/2512.14712v1 Sepsis accounts for nearly 20% of global ICU admissions, yet conventional prediction models often fail to effectively integrate heterogeneous data streams, remaining either siloed by modality or reliant on brittle early fusion. In this work, we present a rigorous architectural comparison between End-to-End Deep Fusion and Context-Aware Stacking for sepsis tasks. We initially hypothesized that a novel Quad-Modal Hierarchical Gated Attention Network -- termed SepsisFusionFormer -- would resolve complex cross-modal interactions between vitals, text, and imaging. However, experiments on MIMIC-IV revealed that SepsisFusionFormer suffered from "attention starvation" in the small antibiotic cohort ($N \approx 2,100$), resulting in overfitting (AUC 0.66). This counterintuitive result informed the design of SepsisLateFusion, a "leaner" Context-Aware Mixture-of-Experts (MoE) architecture. By treating modalities as orthogonal experts -- the "Historian" (Static), the "Monitor" (Temporal), and the "Reader" (NLP) -- and dynamically gating them via a CatBoost meta-learner, we achieved State-of-the-Art (SOTA) performance: 0.915 AUC for prediction 4 hours prior to clinical onset. By calibrating the decision threshold for clinical safety, we reduced missed cases by 48% relative to the default operating point, thus opening a true preventative window for timely intervention over reactive alerts. Furthermore, for the novel prescriptive task of multi-class antibiotic selection, we demonstrate that a Quad-Modal Ensemble achieved the highest performance (0.72 AUC). These models are integrated into SepsisSuite, a deployment-ready Python framework for clinical decision support. SepsisSuite is available for free at: https://github.com/RyanCartularo/SepsisSuite-Info --- ### Promoting Fairness in Information Access within Social Networks Authors: Changan Liu, Xiaotian Zhou, Ahad N. Zehmakan, Zhongzhi Zhang Published: 2025-12-08 Categories: cs.SI, cs.AI arXiv: https://arxiv.org/abs/2512.14711v1 The advent of online social networks has facilitated fast and wide spread of information. However, some users, especially members of minority groups, may be less likely to receive information spreading on the network, due to their disadvantaged network position. We study the optimization problem of adding new connections to a network to enhance fairness in information access among different demographic groups. We provide a concrete formulation of this problem where information access is measured in terms of resistance distance, {offering a new perspective that emphasizes global network structure and multi-path connectivity.} The problem is shown to be NP-hard. We propose a simple greedy algorithm which turns out to output accurate solutions, but its run time is cubic, which makes it undesirable for large networks. As our main technical contribution, we reduce its time complexity to linear, leveraging several novel approximation techniques. In addition to our theoretical findings, we also conduct an extensive set of experiments using both real-world and synthetic datasets. We demonstrate that our linear-time algorithm can produce accurate solutions for networks with millions of nodes. --- ### Autonomous Source Knowledge Selection in Multi-Domain Adaptation Authors: Keqiuyin Li, Jie Lu, Hua Zuo, Guangquan Zhang Published: 2025-12-08 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2512.14710v1 Unsupervised multi-domain adaptation plays a key role in transfer learning by leveraging acquired rich source information from multiple source domains to solve target task from an unlabeled target domain. However, multiple source domains often contain much redundant or unrelated information which can harm transfer performance, especially when in massive-source domain settings. It is urgent to develop effective strategies for identifying and selecting the most transferable knowledge from massive source domains to address the target task. In this paper, we propose a multi-domain adaptation method named \underline{\textit{Auto}}nomous Source Knowledge \underline{\textit{S}}election (AutoS) to autonomosly select source training samples and models, enabling the prediction of target task using more relevant and transferable source information. The proposed method employs a density-driven selection strategy to choose source samples during training and to determine which source models should contribute to target prediction. Simulteneously, a pseudo-label enhancement module built on a pre-trained multimodal modal is employed to mitigate target label noise and improve self-supervision. Experiments on real-world datasets indicate the superiority of the proposed method. --- ### Attention as Binding: A Vector-Symbolic Perspective on Transformer Reasoning Authors: Sahil Rajesh Dhayalkar Published: 2025-12-08 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2512.14709v1 Transformer-based language models display impressive reasoning-like behavior, yet remain brittle on tasks that require stable symbolic manipulation. This paper develops a unified perspective on these phenomena by interpreting self-attention and residual streams as implementing an approximate Vector Symbolic Architecture (VSA). In this view, queries and keys define role spaces, values encode fillers, attention weights perform soft unbinding, and residual connections realize superposition of many bound structures. We use this algebraic lens to relate transformer internals to chain-of-thought traces, program-based reasoning, and memory-augmented tool use, and to explain characteristic failure modes such as variable confusion and inconsistency across logically related prompts. Building on this perspective, we propose VSA-inspired architectural biases, including explicit binding/unbinding heads and hyperdimensional memory layers, and training objectives that promote role-filler separation and robust superposition. Finally, we outline metrics for measuring "VSA-likeness" and logical compositionality, and pose theoretical and architectural open problems. Overall, the paper argues that viewing attention as soft vector-symbolic computation offers a principled route toward more interpretable and logically reliable reasoning systems. --- ### SGEMAS: A Self-Growing Ephemeral Multi-Agent System for Unsupervised Online Anomaly Detection via Entropic Homeostasis Authors: Mustapha Hamdi Published: 2025-12-08 Categories: cs.NE, cs.LG, cs.MA arXiv: https://arxiv.org/abs/2512.14708v1 Current deep learning approaches for physiological signal monitoring suffer from static topologies and constant energy consumption. We introduce SGEMAS (Self-Growing Ephemeral Multi-Agent System), a bio-inspired architecture that treats intelligence as a dynamic thermodynamic process. By coupling a structural plasticity mechanism (agent birth death) to a variational free energy objective, the system naturally evolves to minimize prediction error with extreme sparsity. An ablation study on the MIT-BIH Arrhythmia Database reveals that adding a multi-scale instability index to the agent dynamics significantly improves performance. In a challenging inter-patient, zero-shot setting, the final SGEMAS v3.3 model achieves a mean AUC of 0.570 +- 0.070, outperforming both its simpler variants and a standard autoencoder baseline. This result validates that a physics-based, energy-constrained model can achieve robust unsupervised anomaly detection, offering a promising direction for efficient biomedical AI. --- ### Meta Hierarchical Reinforcement Learning for Scalable Resource Management in O-RAN Authors: Fatemeh Lotfi, Fatemeh Afghah Published: 2025-12-08 Categories: cs.AI, cs.LG, eess.SY arXiv: https://arxiv.org/abs/2512.13715v1 The increasing complexity of modern applications demands wireless networks capable of real time adaptability and efficient resource management. The Open Radio Access Network (O-RAN) architecture, with its RAN Intelligent Controller (RIC) modules, has emerged as a pivotal solution for dynamic resource management and network slicing. While artificial intelligence (AI) driven methods have shown promise, most approaches struggle to maintain performance under unpredictable and highly dynamic conditions. This paper proposes an adaptive Meta Hierarchical Reinforcement Learning (Meta-HRL) framework, inspired by Model Agnostic Meta Learning (MAML), to jointly optimize resource allocation and network slicing in O-RAN. The framework integrates hierarchical control with meta learning to enable both global and local adaptation: the high-level controller allocates resources across slices, while low level agents perform intra slice scheduling. The adaptive meta-update mechanism weights tasks by temporal difference error variance, improving stability and prioritizing complex network scenarios. Theoretical analysis establishes sublinear convergence and regret guarantees for the two-level learning process. Simulation results demonstrate a 19.8% improvement in network management efficiency compared with baseline RL and meta-RL approaches, along with faster adaptation and higher QoS satisfaction across eMBB, URLLC, and mMTC slices. Additional ablation and scalability studies confirm the method's robustness, achieving up to 40% faster adaptation and consistent fairness, latency, and throughput performance as network scale increases. --- ### AI-Powered Annotation Pipelines for Stabilizing Large Language Models: A Human-AI Synergy Approach Authors: Gangesh Pathak, Prasanna Kumar Published: 2025-12-08 Categories: cs.AI arXiv: https://arxiv.org/abs/2512.13714v1 LLM implementations are failing in highly regulated industries owing to instability issues, inconsistent reasoning, hallucinations and performance variability, especially in workflows. These reliability issues restrict safe use of LLM in areas that need the precision of facts and consistent behavior (Aiyappa et al., 2023). The current methods of stabilization, such as, reinforcement learning with human feedback (RLHF) and supervised fine-tuning, offer quantifiable improvements but are expensive and based on the intensive annotation of humans, thus being not easily scaled in a sustainable way (Dong et al., 2023; Retzlaff et al., 2024). This paper presents an AI-based annotation pipeline that systematically identifies, labels, and fixes for instability patterns on LLM output. Our human-AI synergy method combines the models of automated weak supervision and confidence-based annotation with the target human validation to guarantee the reliability and moral uprightness of feedback information (Cabitza et al., 2023; Jiang et al., 2023). The semantic consistency, factual correctness, and logical coherence categories of stability-specific annotation are introduced into our framework, allowing the continuous calibration of models and the enhancement of their robustness based on the feedback loops (Honovich et al., 2021; Nan et al., 2021). --- ### An Experience Report on a Pedagogically Controlled, Curriculum-Constrained AI Tutor for SE Education Authors: Lucia Happe, Dominik Fuchß, Luca Hüttner, Kai Marquardt, Anne Koziolek Published: 2025-12-08 Categories: cs.CY, cs.AI, cs.HC, cs.SE arXiv: https://arxiv.org/abs/2512.11882v1 The integration of artificial intelligence (AI) into education continues to evoke both promise and skepticism. While past waves of technological optimism often fell short, recent advances in large language models (LLMs) have revived the vision of scalable, individualized tutoring. This paper presents the design and pilot evaluation of RockStartIT Tutor, an AI-powered assistant developed for a digital programming and computational thinking course within the RockStartIT initiative. Powered by GPT-4 via OpenAI's Assistant API, the tutor employs a novel prompting strategy and a modular, semantically tagged knowledge base to deliver context-aware, personalized, and curriculum-constrained support for secondary school students. We evaluated the system using the Technology Acceptance Model (TAM) with 13 students and teachers. Learners appreciated the low-stakes environment for asking questions and receiving scaffolded guidance. Educators emphasized the system's potential to reduce cognitive load during independent tasks and complement classroom teaching. Key challenges include prototype limitations, a small sample size, and the need for long-term studies with the target age group. Our findings highlight a pragmatic approach to AI integration that requires no model training, using structure and prompts to shape behavior. We position AI tutors not as teacher replacements but as enabling tools that extend feedback access, foster inquiry, and support what schools do best: help students learn. --- ### Understanding Structural Representation in Foundation Models for Polymers Authors: Nathaniel H. Park, Eduardo Soares, Victor Y. Shirasuna, Tiffany J. Callahan, Sara Capponi, Emilio Vital Brazil Published: 2025-12-08 Categories: cond-mat.soft, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2512.11881v2 From the relative scarcity of training data to the lack of standardized benchmarks, the creation of effective foundation models for polymers faces significant and multi-faceted challenges. At the core, many of these issues are tied directly to the structural representation of polymers. Here, we present a chemical language foundation model built on using a SMILES-based polymer graph representation (CPG) that incorporates polymer architectural features and connectivity that are often missing in other line notations. This foundation model exhibited excellent performance on 30 different polymer property benchmark datasets. Critical evaluation of the developed representation against other variations in control experiments reveals this approach to be a robust method of representing polymers in language-based foundation models. These experiments also reveal a strong invariance of structural representations to small perturbations, with many variations of structural representation exceeding or equaling state-of-the-art (SOTA) performance. Surprisingly, SMILES representations which are chemically or semantically invalid also provided near or SOTA performance in several instances--underscoring an unexamined blind spot in the development of chemistry language models. Examination of error sources and attention maps for the evaluated structural representations corroborate the findings of the control experiments, highlighting the ability of the model to interpolate SMILES sequence space in a manner that is loosely congruent to chemical and architectural space for polymers. Overall, this work highlights the surprising robustness of chemistry language models to structural representation perturbations and identifies the conditions under which CPG representation provides meaningful advantages. --- ### Shakespeare, Entropy and Educated Monkeys Authors: Ioannis Kontoyiannis Published: 2025-12-08 Categories: math.HO, cs.CL, cs.IT arXiv: https://arxiv.org/abs/2512.11880v1 It has often been said, correctly, that a monkey forever randomly typing on a keyboard would eventually produce the complete works of William Shakespeare. Almost just as often it has been pointed out that this "eventually" is well beyond any conceivably relevant time frame. We point out that an educated monkey that still types at random but is constrained to only write "statistically typical" text, would produce any given passage in a much shorter time. Information theory gives a very simple way to estimate that time. For example, Shakespeare's phrase, Better three hours too soon than a minute too late, from The Merry Wives of Windsor, would take the educated monkey only 73 thousand years to produce, compared to the beyond-astronomical $2.7 \times 10^{63}$ years for the randomly typing one. Despite the obvious improvement, it would still take the educated monkey an unimaginably long $10^{42,277}$ years to produce all of Hamlet. --- ### Generalization of Long-Range Machine Learning Potentials in Complex Chemical Spaces Authors: Michal Sanocki, Julija Zavadlav Published: 2025-12-08 Categories: physics.chem-ph, cs.LG arXiv: https://arxiv.org/abs/2512.10989v2 The vastness of chemical space makes generalization a central challenge in the development of machine learning interatomic potentials (MLIPs). While MLIPs could enable large-scale atomistic simulations with near-quantum accuracy, their usefulness is often limited by poor transferability to out-of-distribution samples. Here, we systematically evaluate different MLIP architectures with long-range corrections across diverse chemical spaces and show that such schemes are essential, not only for improving in-distribution performance but, more importantly, for enabling significant gains in transferability to unseen regions of chemical space. To enable a more rigorous benchmarking, we introduce biased train-test splitting strategies, which explicitly test the model performance in significantly different regions of chemical space. Together, our findings highlight the importance of long-range modeling for achieving generalizable MLIPs and provide a framework for diagnosing systematic failures across chemical space. Although we demonstrate our methodology on metal-organic frameworks, it is broadly applicable to other materials, offering insights into the design of more robust and transferable MLIPs. --- ### HGC-Herd: Efficient Heterogeneous Graph Condensation via Representative Node Herding Authors: Fuyan Ou, Siqi Ai, Yulin Hu Published: 2025-12-08 Categories: cs.LG, cs.IR arXiv: https://arxiv.org/abs/2512.09947v1 Heterogeneous graph neural networks (HGNNs) have demonstrated strong capability in modeling complex semantics across multi-type nodes and relations. However, their scalability to large-scale graphs remains challenging due to structural redundancy and high-dimensional node features. Existing graph condensation approaches, such as GCond, are primarily developed for homogeneous graphs and rely on gradient matching, resulting in considerable computational, memory, and optimization overhead. We propose HGC-Herd, a training-free condensation framework that generates compact yet informative heterogeneous graphs while maintaining both semantic and structural fidelity. HGC-Herd integrates lightweight feature propagation to encode multi-hop relational context and employs a class-wise herding mechanism to identify representative nodes per class, producing balanced and discriminative subsets for downstream learning tasks. Extensive experiments on ACM, DBLP, and Freebase validate that HGC-Herd attains comparable or superior accuracy to full-graph training while markedly reducing both runtime and memory consumption. These results underscore its practical value for efficient and scalable heterogeneous graph representation learning. --- ### Demo: Generative AI helps Radiotherapy Planning with User Preference Authors: Riqiang Gao, Simon Arberet, Martin Kraus, Han Liu, Wilko FAR Verbakel, Dorin Comaniciu, Florin-Cristian Ghesu, Ali Kamen Published: 2025-12-08 Categories: cs.CV, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2512.08996v2 Radiotherapy planning is a highly complex process that often varies significantly across institutions and individual planners. Most existing deep learning approaches for 3D dose prediction rely on reference plans as ground truth during training, which can inadvertently bias models toward specific planning styles or institutional preferences. In this study, we introduce a novel generative model that predicts 3D dose distributions based solely on user-defined preference flavors. These customizable preferences enable planners to prioritize specific trade-offs between organs-at-risk (OARs) and planning target volumes (PTVs), offering greater flexibility and personalization. Designed for seamless integration with clinical treatment planning systems, our approach assists users in generating high-quality plans efficiently. Comparative evaluations demonstrate that our method can surpasses the Varian RapidPlan model in both adaptability and plan quality in some scenarios. --- ### Enhanced Chest Disease Classification Using an Improved CheXNet Framework with EfficientNetV2-M and Optimization-Driven Learning Authors: Ali M. Bahram, Saman Muhammad Omer, Hardi M. Mohammed, Sirwan Abdolwahed Aula Published: 2025-12-08 Categories: eess.IV, cs.AI, cs.CV, cs.LG arXiv: https://arxiv.org/abs/2512.08992v1 The interpretation of Chest X-ray is an important diagnostic issue in clinical practice and especially in the resource-limited setting where the shortage of radiologists plays a role in delayed diagnosis and poor patient outcomes. Although the original CheXNet architecture has shown potential in automated analysis of chest radiographs, DenseNet-121 backbone is computationally inefficient and poorly single-label classifier. To eliminate such shortcomings, we suggest a better classification framework of chest disease that relies on EfficientNetV2-M and incorporates superior training approaches such as Automatic Mixed Precision training, AdamW, Cosine Annealing learning rate scheduling, and Exponential Moving Average regularization. We prepared a dataset of 18,080 chest X-ray images of three source materials of high authority and representing five key clinically significant disease categories which included Cardiomegaly, COVID-19, Normal, Pneumonia, and Tuberculosis. To achieve statistical reliability and reproducibility, nine independent experimental runs were run. The suggested architecture showed significant gains with mean test accuracy of 96.45 percent compared to 95.30 percent at baseline (p less than 0.001) and macro-averaged F1-score increased to 91.08 percent (p less than 0.001). Critical infectious diseases showed near-perfect classification performance with COVID-19 detection having 99.95 percent accuracy and Tuberculosis detection having 99.97 percent accuracy. Although 6.8 times more parameters are included, the training time was reduced by 11.4 percent and performance stability was increased by 22.7 percent. This framework presents itself as a decision-support tool that can be used to respond to a pandemic, screen tuberculosis, and assess thoracic disease regularly in various healthcare facilities. --- ### Deterministic World Models for Closed-loop Reachability Analysis of End-to-End Vision-based Control Authors: Yuang Geng, Zhongzheng Zhang, Chengzhen Jiang, Yanru Li, Xinyang Wang, Zhuoyang Zhou, Hoang-Dung Tran, Ivan Ruchkin Published: 2025-12-08 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2512.08991v3 End-to-end image controllers that map raw camera frames directly to control actions are increasingly deployed in safety-critical systems. However, formally verifying their closed-loop behavior remains an open challenge because cameras produce high-dimensional images whose generation cannot easily be described in a closed mathematical form. We propose a Deterministic World Model (DWM), a latent-free neural decoder that maps physical states (e.g., position and velocity) directly to synthetic camera images, enabling closed-loop reachability analysis without the overapproximation caused by stochastic latent variables. The DWM is trained with a novel dual loss combining saliency-map reconstruction and a control-consistent term that preserves behavioral consistency with the real controller. We integrate the DWM into closed-loop reachability analysis and apply conformal prediction to inflate the reachable sets by a distribution-free trajectory-tube deviation bound, transferring the surrogate guarantee to the real system with high probability. Experiments on a CARLA braking system and three Gym benchmarks (CartPole, MountainCar, Pendulum) show that the DWM produces substantially tighter reachable tubes than a cGAN and trajectory predictor baselines while meeting the target coverage after conformal inflation. --- ### Agreement Disagreement Guided Knowledge Transfer for Cross-Scene Hyperspectral Imaging Authors: Lu Huo, Haimin Zhang, Min Xu Published: 2025-12-08 Categories: eess.IV, cs.CV arXiv: https://arxiv.org/abs/2512.08990v1 Knowledge transfer plays a crucial role in cross-scene hyperspectral imaging (HSI). However, existing studies often overlook the challenges of gradient conflicts and dominant gradients that arise during the optimization of shared parameters. Moreover, many current approaches fail to simultaneously capture both agreement and disagreement information, relying only on a limited shared subset of target features and consequently missing the rich, diverse patterns present in the target scene. To address these issues, we propose an Agreement Disagreement Guided Knowledge Transfer (ADGKT) framework that integrates both mechanisms to enhance cross-scene transfer. The agreement component includes GradVac, which aligns gradient directions to mitigate conflicts between source and target domains, and LogitNorm, which regulates logit magnitudes to prevent domination by a single gradient source. The disagreement component consists of a Disagreement Restriction (DiR) and an ensemble strategy, which capture diverse predictive target features and mitigate the loss of critical target information. Extensive experiments demonstrate the effectiveness and superiority of the proposed method in achieving robust and balanced knowledge transfer across heterogeneous HSI scenes. --- ### Enhancing Knowledge Transfer in Hyperspectral Image Classification via Cross-scene Knowledge Integration Authors: Lu Huo, Wenjian Huang, Jianguo Zhang, Min Xu, Haimin Zhang Published: 2025-12-08 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.08989v1 Knowledge transfer has strong potential to improve hyperspectral image (HSI) classification, yet two inherent challenges fundamentally restrict effective cross-domain transfer: spectral variations caused by different sensors and semantic inconsistencies across heterogeneous scenes. Existing methods are limited by transfer settings that assume homogeneous domains or heterogeneous scenarios with only co-occurring categories. When label spaces do not overlap, they further rely on complete source-domain coverage and therefore overlook critical target-private information. To overcome these limitations and enable knowledge transfer in fully heterogeneous settings, we propose Cross-scene Knowledge Integration (CKI), a framework that explicitly incorporates target-private knowledge during transfer. CKI includes: (1) Alignment of Spectral Characteristics (ASC) to reduce spectral discrepancies through domain-agnostic projection; (2) Cross-scene Knowledge Sharing Preference (CKSP), which resolves semantic mismatch via a Source Similarity Mechanism (SSM); and (3) Complementary Information Integration (CII) to maximize the use of target-specific complementary cues. Extensive experiments verify that CKI achieves state-of-the-art performance with strong stability across diverse cross-scene HSI scenarios. --- ### Balanced Accuracy: The Right Metric for Evaluating LLM Judges -- Explained through Youden's J statistic Authors: Stephane Collot, Colin Fraser, Justin Zhao, William F. Shen, Timon Willi, Ilias Leontiadis Published: 2025-12-08 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2512.08121v2 Rigorous evaluation of large language models (LLMs) relies on comparing models by the prevalence of desirable or undesirable behaviors, such as task pass rates or policy violations. These prevalence estimates are produced by a classifier, either an LLM-as-a-judge or human annotators, making the choice of classifier central to trustworthy evaluation. Common metrics used for this choice, such as Accuracy, Precision, and F1, are sensitive to class imbalance and to arbitrary choices of positive class, and can favor judges that distort prevalence estimates. We show that Youden's $J$ statistic is theoretically aligned with choosing the best judge to compare models, and that Balanced Accuracy is an equivalent linear transformation of $J$. Through both analytical arguments and empirical examples and simulations, we demonstrate how selecting judges using Balanced Accuracy leads to better, more robust classifier selection. --- ### Scalable Offline Model-Based RL with Action Chunks Authors: Kwanyoung Park, Seohong Park, Youngwoon Lee, Sergey Levine Published: 2025-12-08 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2512.08108v1 In this paper, we study whether model-based reinforcement learning (RL), in particular model-based value expansion, can provide a scalable recipe for tackling complex, long-horizon tasks in offline RL. Model-based value expansion fits an on-policy value function using length-n imaginary rollouts generated by the current policy and a learned dynamics model. While larger n reduces bias in value bootstrapping, it amplifies accumulated model errors over long horizons, degrading future predictions. We address this trade-off with an \emph{action-chunk} model that predicts a future state from a sequence of actions (an "action chunk") instead of a single action, which reduces compounding errors. In addition, instead of directly training a policy to maximize rewards, we employ rejection sampling from an expressive behavioral action-chunk policy, which prevents model exploitation from out-of-distribution actions. We call this recipe \textbf{Model-Based RL with Action Chunks (MAC)}. Through experiments on highly challenging tasks with large-scale datasets of up to 100M transitions, we show that MAC achieves the best performance among offline model-based RL algorithms, especially on challenging long-horizon tasks. --- ### Generalizations of the Normalized Radon Cumulative Distribution Transform for Limited Data Recognition Authors: Matthias Beckmann, Robert Beinert, Jonas Bresch Published: 2025-12-08 Categories: math.NA, cs.CV, cs.IT arXiv: https://arxiv.org/abs/2512.08099v2 The Radon cumulative distribution transform (R-CDT) exploits one-dimensional Wasserstein transport and the Radon transform to represent prominent features in images. It is closely related to the sliced Wasserstein distance and facilitates classification tasks, especially in the small data regime, like the recognition of watermarks in filigranology. Here, a typical issue is that the given data may be subject to affine transformations caused by the measuring process. To make the R-CDT invariant under arbitrary affine transformations, a two-step normalization of the R-CDT has been proposed in our earlier works. The aim of this paper is twofold. First, we propose a family of generalized normalizations to enhance flexibility for applications. Second, we study multi-dimensional and non-Euclidean settings by making use of generalized Radon transforms. We prove that our novel feature representations are invariant under certain transformations and allow for linear separation in feature space. Our theoretical results are supported by numerical experiments based on 2d images, 3d shapes and 3d rotation matrices, showing near perfect classification accuracies and clustering results. --- ### Segment, Embed, and Align: A Universal Recipe for Aligning Subtitles to Signing Authors: Zifan Jiang, Youngjoon Jang, Liliane Momeni, Gül Varol, Sarah Ebling, Andrew Zisserman Published: 2025-12-08 Categories: cs.CL arXiv: https://arxiv.org/abs/2512.08094v2 The goal of this work is to develop a universal approach for aligning subtitles (i.e., spoken language text with corresponding timestamps) to continuous sign language videos. Prior approaches typically rely on end-to-end training tied to a specific language or dataset, which limits their generality. In contrast, our method Segment, Embed, and Align (SEA) provides a single framework that works across multiple languages and domains. SEA leverages two pretrained models: the first to segment a video frame sequence into individual signs and the second to embed the video clip of each sign into a shared latent space with text. Alignment is subsequently performed with a lightweight dynamic programming procedure that runs efficiently on CPUs within a minute, even for hour-long episodes. SEA is flexible and can adapt to a wide range of scenarios, utilizing resources from small lexicons to large continuous corpora. Experiments on four sign language datasets demonstrate state-of-the-art alignment performance, highlighting the potential of SEA to generate high-quality parallel data for advancing sign language processing. SEA's code and models are openly available. --- ### Training LLMs for Honesty via Confessions Authors: Manas Joglekar, Jeremy Chen, Gabriel Wu, Jason Yosinski, Jasmine Wang, Boaz Barak, Amelia Glaese Published: 2025-12-08 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2512.08093v2 Large language models (LLMs) can be dishonest when reporting on their actions and beliefs -- for example, they may overstate their confidence in factual claims or cover up evidence of covert actions. Such dishonesty may arise due to the effects of reinforcement learning (RL), where challenges with reward shaping can result in a training process that inadvertently incentivizes the model to lie or misrepresent its actions. In this work we propose a method for eliciting an honest expression of an LLM's shortcomings via a self-reported *confession*. A confession is an output, provided upon request after a model's original answer, that is meant to serve as a full account of the model's compliance with the letter and spirit of its policies and instructions. The reward assigned to a confession during training is solely based on its honesty, and does not impact positively or negatively the main answer's reward. As long as the "path of least resistance" for maximizing confession reward is to surface misbehavior rather than covering it up, this incentivizes models to be honest in their confessions. Our findings provide some justification this empirical assumption, especially in the case of egregious model misbehavior. To demonstrate the viability of our approach, we train GPT-5-Thinking to produce confessions, and we evaluate its honesty in out-of-distribution scenarios measuring hallucination, instruction following, scheming, and reward hacking. We find that when the model lies or omits shortcomings in its "main" answer, it often confesses to these behaviors honestly, and this confession honesty modestly improves with training. Confessions can enable a number of inference-time interventions including monitoring, rejection sampling, and surfacing issues to the user. --- ### Complexity of One-Dimensional ReLU DNNs Authors: Jonathan Kogan, Hayden Jananthan, Jeremy Kepner Published: 2025-12-08 Categories: cs.LG, stat.ML arXiv: https://arxiv.org/abs/2512.08091v1 We study the expressivity of one-dimensional (1D) ReLU deep neural networks through the lens of their linear regions. For randomly initialized, fully connected 1D ReLU networks (He scaling with nonzero bias) in the infinite-width limit, we prove that the expected number of linear regions grows as $\sum_{i = 1}^L n_i + \mathop{o}\left(\sum_{i = 1}^L{n_i}\right) + 1$, where $n_\ell$ denotes the number of neurons in the $\ell$-th hidden layer. We also propose a function-adaptive notion of sparsity that compares the expected regions used by the network to the minimal number needed to approximate a target within a fixed tolerance. --- ### Adaptation of Embedding Models to Financial Filings via LLM Distillation Authors: Eliot Brenner, Dominic Seyler, Manjunath Hegde, Andrei Simion, Koustuv Dasgupta, Bing Xiang Published: 2025-12-08 Categories: cs.CL arXiv: https://arxiv.org/abs/2512.08088v1 Despite advances in generative large language models (LLMs), practical application of specialized conversational AI agents remains constrained by computation costs, latency requirements, and the need for precise domain-specific relevance measures. While existing embedding models address the first two constraints, they underperform on information retrieval in specialized domains like finance. This paper introduces a scalable pipeline that trains specialized models from an unlabeled corpus using a general purpose retrieval embedding model as foundation. Our method yields an average of 27.7% improvement in MRR$\texttt{@}$5, 44.6% improvement in mean DCG$\texttt{@}$5 across 14 financial filing types measured over 21,800 query-document pairs, and improved NDCG on 3 of 4 document classes in FinanceBench. We adapt retrieval embeddings (bi-encoder) for RAG, not LLM generators, using LLM-judged relevance to distill domain knowledge into a compact retriever. There are prior works which pair synthetically generated queries with real passages to directly fine-tune the retrieval model. Our pipeline differs from these by introducing interaction between student and teacher models that interleaves retrieval-based mining of hard positive/negative examples from the unlabeled corpus with iterative retraining of the student model's weights using these examples. Each retrieval iteration uses the refined student model to mine the corpus for progressively harder training examples for the subsequent training iteration. The methodology provides a cost-effective solution to bridging the gap between general-purpose models and specialized domains without requiring labor-intensive human annotation. --- ### Short-Context Dominance: How Much Local Context Natural Language Actually Needs? Authors: Vala Vakilian, Zimeng Wang, Ankit Singh Rawat, Christos Thrampoulidis Published: 2025-12-08 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2512.08082v2 We investigate the short-context dominance hypothesis: that for most sequences, a small local prefix suffices to predict their next tokens. Using large language models as statistical oracles, we measure the minimum context length (MCL) needed to reproduce accurate full-context predictions across datasets with sequences of varying lengths. For sequences with 1-7k tokens from long-context documents, we consistently find that 75-80% require only the last 96 tokens at most. Given the dominance of short-context tokens, we then ask whether it is possible to detect challenging long-context sequences for which a short local prefix does not suffice for prediction. We introduce a practical proxy to MCL, called Distributionally Aware MCL (DaMCL), that does not require knowledge of the actual next-token and is compatible with sampling strategies beyond greedy decoding. Our experiments validate that simple thresholding of the metric defining DaMCL achieves high performance in detecting long vs. short context sequences. Finally, to counter the bias that short-context dominance induces in LLM output distributions, we develop an intuitive decoding algorithm that leverages our detector to identify and boost tokens that are long-range-relevant. Across Q&A tasks and model architectures, we confirm that mitigating the bias improves performance. --- ### Unveiling Latent Knowledge in Chemistry Language Models through Sparse Autoencoders Authors: Jaron Cohen, Alexander G. Hasson, Sara Tanovic Published: 2025-12-08 Categories: cs.LG, physics.chem-ph arXiv: https://arxiv.org/abs/2512.08077v1 Since the advent of machine learning, interpretability has remained a persistent challenge, becoming increasingly urgent as generative models support high-stakes applications in drug and material discovery. Recent advances in large language model (LLM) architectures have yielded chemistry language models (CLMs) with impressive capabilities in molecular property prediction and molecular generation. However, how these models internally represent chemical knowledge remains poorly understood. In this work, we extend sparse autoencoder techniques to uncover and examine interpretable features within CLMs. Applying our methodology to the Foundation Models for Materials (FM4M) SMI-TED chemistry foundation model, we extract semantically meaningful latent features and analyse their activation patterns across diverse molecular datasets. Our findings reveal that these models encode a rich landscape of chemical concepts. We identify correlations between specific latent features and distinct domains of chemical knowledge, including structural motifs, physicochemical properties, and pharmacological drug classes. Our approach provides a generalisable framework for uncovering latent knowledge in chemistry-focused AI systems. This work has implications for both foundational understanding and practical deployment; with the potential to accelerate computational chemistry research. --- ### Identification of Deforestation Areas in the Amazon Rainforest Using Change Detection Models Authors: Christian Massao Konishi, Helio Pedrini Published: 2025-12-08 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.08075v1 The preservation of the Amazon Rainforest is one of the global priorities in combating climate change, protecting biodiversity, and safeguarding indigenous cultures. The Satellite-based Monitoring Project of Deforestation in the Brazilian Legal Amazon (PRODES), a project of the National Institute for Space Research (INPE), stands out as a fundamental initiative in this effort, annually monitoring deforested areas not only in the Amazon but also in other Brazilian biomes. Recently, machine learning models have been developed using PRODES data to support this effort through the comparative analysis of multitemporal satellite images, treating deforestation detection as a change detection problem. However, existing approaches present significant limitations: models evaluated in the literature still show unsatisfactory effectiveness, many do not incorporate modern architectures, such as those based on self-attention mechanisms, and there is a lack of methodological standardization that allows direct comparisons between different studies. In this work, we address these gaps by evaluating various change detection models in a unified dataset, including fully convolutional models and networks incorporating self-attention mechanisms based on Transformers. We investigate the impact of different pre- and post-processing techniques, such as filtering deforested areas predicted by the models based on the size of connected components, texture replacement, and image enhancements; we demonstrate that such approaches can significantly improve individual model effectiveness. Additionally, we test different strategies for combining the evaluated models to achieve results superior to those obtained individually, reaching an F1-score of 80.41%, a value comparable to other recent works in the literature. --- ### CAMO: Causality-Guided Adversarial Multimodal Domain Generalization for Crisis Classification Authors: Pingchuan Ma, Chengshuai Zhao, Bohan Jiang, Saketh Vishnubhatla, Ujun Jeong, Alimohammad Beigi, Adrienne Raglin, Huan Liu Published: 2025-12-08 Categories: cs.LG arXiv: https://arxiv.org/abs/2512.08071v1 Crisis classification in social media aims to extract actionable disaster-related information from multimodal posts, which is a crucial task for enhancing situational awareness and facilitating timely emergency responses. However, the wide variation in crisis types makes achieving generalizable performance across unseen disasters a persistent challenge. Existing approaches primarily leverage deep learning to fuse textual and visual cues for crisis classification, achieving numerically plausible results under in-domain settings. However, they exhibit poor generalization across unseen crisis types because they 1. do not disentangle spurious and causal features, resulting in performance degradation under domain shift, and 2. fail to align heterogeneous modality representations within a shared space, which hinders the direct adaptation of established single-modality domain generalization (DG) techniques to the multimodal setting. To address these issues, we introduce a causality-guided multimodal domain generalization (MMDG) framework that combines adversarial disentanglement with unified representation learning for crisis classification. The adversarial objective encourages the model to disentangle and focus on domain-invariant causal features, leading to more generalizable classifications grounded in stable causal mechanisms. The unified representation aligns features from different modalities within a shared latent space, enabling single-modality DG strategies to be seamlessly extended to multimodal learning. Experiments on the different datasets demonstrate that our approach achieves the best performance in unseen disaster scenarios. --- ### Deep Kernel Aalen-Johansen Estimator: An Interpretable and Flexible Neural Net Framework for Competing Risks Authors: Xiaobin Shen, George H. Chen Published: 2025-12-08 Categories: cs.LG arXiv: https://arxiv.org/abs/2512.08063v1 We propose an interpretable deep competing risks model called the Deep Kernel Aalen-Johansen (DKAJ) estimator, which generalizes the classical Aalen-Johansen nonparametric estimate of cumulative incidence functions (CIFs). Each data point (e.g., patient) is represented as a weighted combination of clusters. If a data point has nonzero weight only for one cluster, then its predicted CIFs correspond to those of the classical Aalen-Johansen estimator restricted to data points from that cluster. These weights come from an automatically learned kernel function that measures how similar any two data points are. On four standard competing risks datasets, we show that DKAJ is competitive with state-of-the-art baselines while being able to provide visualizations to assist model interpretation. --- ### LUNA: Linear Universal Neural Attention with Generalization Guarantees Authors: Ashkan Shahbazi, Ping He, Ali Abbasi, Yikun Bai, Xinran Liu, Elaheh Akbari, Darian Salehi, Navid NaderiAlizadeh, Soheil Kolouri Published: 2025-12-08 Categories: cs.LG, stat.ML arXiv: https://arxiv.org/abs/2512.08061v1 Scaling attention faces a critical bottleneck: the $\mathcal{O}(n^2)$ quadratic computational cost of softmax attention, which limits its application in long-sequence domains. While linear attention mechanisms reduce this cost to $\mathcal{O}(n)$, they typically rely on fixed random feature maps, such as random Fourier features or hand-crafted functions. This reliance on static, data-agnostic kernels creates a fundamental trade-off, forcing practitioners to sacrifice significant model accuracy for computational efficiency. We introduce \textsc{LUNA}, a kernelized linear attention mechanism that eliminates this trade-off, retaining linear cost while matching and surpassing the accuracy of quadratic attention. \textsc{LUNA} is built on the key insight that the kernel feature map itself should be learned rather than fixed a priori. By parameterizing the kernel, \textsc{LUNA} learns a feature basis tailored to the specific data and task, overcoming the expressive limitations of fixed-feature methods. \textsc{Luna} implements this with a learnable feature map that induces a positive-definite kernel and admits a streaming form, yielding linear time and memory scaling in the sequence length. Empirical evaluations validate our approach across diverse settings. On the Long Range Arena (LRA), \textsc{Luna} achieves state-of-the-art average accuracy among efficient Transformers under compute parity, using the same parameter count, training steps, and approximate FLOPs. \textsc{Luna} also excels at post-hoc conversion: replacing softmax in fine-tuned BERT and ViT-B/16 checkpoints and briefly fine-tuning recovers most of the original performance, substantially outperforming fixed linearizations. --- ### Large Language Models for Education and Research: An Empirical and User Survey-based Analysis Authors: Md Mostafizer Rahman, Ariful Islam Shiplu, Md Faizul Ibne Amin, Yutaka Watanobe, Lu Peng Published: 2025-12-08 Categories: cs.AI arXiv: https://arxiv.org/abs/2512.08057v1 Pretrained Large Language Models (LLMs) have achieved remarkable success across diverse domains, with education and research emerging as particularly impactful areas. Among current state-of-the-art LLMs, ChatGPT and DeepSeek exhibit strong capabilities in mathematics, science, medicine, literature, and programming. In this study, we present a comprehensive evaluation of these two LLMs through background technology analysis, empirical experiments, and a real-world user survey. The evaluation explores trade-offs among model accuracy, computational efficiency, and user experience in educational and research affairs. We benchmarked these LLMs performance in text generation, programming, and specialized problem-solving. Experimental results show that ChatGPT excels in general language understanding and text generation, while DeepSeek demonstrates superior performance in programming tasks due to its efficiency-focused design. Moreover, both models deliver medically accurate diagnostic outputs and effectively solve complex mathematical problems. Complementing these quantitative findings, a survey of students, educators, and researchers highlights the practical benefits and limitations of these models, offering deeper insights into their role in advancing education and research. --- ### Fairness-aware PageRank via Edge Reweighting Authors: Honglian Wang, Haoyun Chen, Aristides Gionis Published: 2025-12-08 Categories: cs.SI, cs.LG arXiv: https://arxiv.org/abs/2512.08055v2 Link-analysis algorithms, such as PageRank, are instrumental in understanding the structural dynamics of networks by evaluating the importance of individual vertices based on their connectivity. Recently, with the rising importance of responsible AI, the question of fairness in link-analysis algorithms has gained traction. In this paper, we present a new approach for incorporating group fairness into the PageRank algorithm by reweighting the transition probabilities in the underlying transition matrix. We formulate the problem of achieving fair PageRank by seeking to minimize the fairness loss, which is the difference between the original group-wise PageRank distribution and a target PageRank distribution. We further define a group-adapted fairness notion, which accounts for group homophily by considering random walks with group-biased restart for each group. Since the fairness loss is non-convex, we propose an efficient projected gradient-descent method for computing locally-optimal edge weights. Unlike earlier approaches, we do not recommend adding new edges to the network, nor do we adjust the restart vector. Instead, we keep the topology of the underlying network unchanged and only modify the relative importance of existing edges. We empirically compare our approach with state-of-the-art baselines and demonstrate the efficacy of our method, where very small changes in the transition matrix lead to significant improvement in the fairness of the PageRank algorithm. --- ### An Introduction to Deep Reinforcement and Imitation Learning Authors: Pedro Santana Published: 2025-12-08 Categories: cs.RO, cs.LG arXiv: https://arxiv.org/abs/2512.08052v3 Embodied agents, such as robots and virtual characters, must continuously select actions to execute tasks effectively, solving complex sequential decision-making problems. Given the difficulty of designing such controllers manually, learning-based approaches have emerged as promising alternatives, most notably Deep Reinforcement Learning (DRL) and Deep Imitation Learning (DIL). DRL leverages reward signals to optimize behavior, while DIL uses expert demonstrations to guide learning. This document introduces DRL and DIL in the context of embodied agents, adopting a concise, depth-first approach to the literature. It is self-contained, presenting all necessary mathematical and machine learning concepts as they are needed. It is not intended as a survey of the field; rather, it focuses on a small set of foundational algorithms and techniques, prioritizing in-depth understanding over broad coverage. The material ranges from Markov Decision Processes to REINFORCE and Proximal Policy Optimization (PPO) for DRL, and from Behavioral Cloning to Dataset Aggregation (DAgger) and Generative Adversarial Imitation Learning (GAIL) for DIL. --- ### Family Matters: A Systematic Study of Spatial vs. Frequency Masking for Continual Test-Time Adaptation Authors: Chandler Timm C. Doloriel, Yunbei Zhang, Yeonguk Yu, Taki Hasan Rafi, Muhammad salman siddiqui, Tor Kristian Stevik, Fadi Al Machot, Kristian Hovde Liland, Habib Ullah Published: 2025-12-08 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.08048v4 Recent continual test-time adaptation (CTTA) methods adopt masked image modeling to stabilize learning under distribution shift, yet each treats its masking family F as a fixed design choice and innovates exclusively along the selection strategy S, leaving the family axis underexplored. We present a systematic empirical study that isolates this axis. Using a controlled CTTA instantiation, Mask to Adapt (M2A), that fixes S=random and standard losses, we vary only F across spatial (patch, pixel) and frequency (all-band, low-band, high-band) families while keeping every other component identical. The study's contributions are the design guidance it extracts for the CTTA settings we evaluated: (1) the masking family determines whether adaptation compounds useful structure or compounds errors, on patch-tokenized architectures, spatial masking accumulates stable representations over long streams while frequency masking collapses catastrophically. We characterize this instability through a structural-preservation account, where spatial coherence maintains the broad-spectrum redundancy needed to avoid terminally overlapping with a corruption's spectral signature; (2) the optimal family depends on architecture-task alignment, on CNNs, whose overlapping receptive fields dilute patch occlusion, the family gap vanishes, whereas on fine-grained tasks with global cues and large-capacity ViTs, frequency masking becomes competitive. In confounded system-level comparisons, where baselines also differ in losses and auxiliary components, M2A's random selection performs comparably to heuristic strategies, though we treat this observation as suggestive context rather than a controlled quantification of S's relative importance. --- ### Towards Sustainable Universal Deepfake Detection with Frequency-Domain Masking Authors: Chandler Timm C. Doloriel, Habib Ullah, Kristian Hovde Liland, Fadi Al Machot, Ngai-Man Cheung Published: 2025-12-08 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.08042v3 Universal deepfake detection aims to identify AI-generated images across a broad range of generative models, including unseen ones. This requires robust generalization to new and unseen deepfakes, which emerge frequently, while minimizing computational overhead to enable large-scale deepfake screening, a critical objective in the era of Green AI. In this work, we explore frequency-domain masking as a training strategy for deepfake detectors. Unlike traditional methods that rely heavily on spatial features or large-scale pretrained models, our approach introduces random masking and geometric transformations, with a focus on frequency masking due to its superior generalization properties. We demonstrate that frequency masking not only enhances detection accuracy across diverse generators but also maintains performance under significant model pruning, offering a scalable and resource-conscious solution. Our method achieves state-of-the-art generalization on GAN- and diffusion-generated image datasets and exhibits consistent robustness under structured pruning. These results highlight the potential of frequency-based masking as a practical step toward sustainable and generalizable deepfake detection. Code and models are available at https://github.com/chandlerbing65nm/FakeImageDetection. --- ### Lost in Translation, Found in Embeddings: Sign Language Translation and Alignment Authors: Youngjoon Jang, Liliane Momeni, Zifan Jiang, Joon Son Chung, Gül Varol, Andrew Zisserman Published: 2025-12-08 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.08040v1 Our aim is to develop a unified model for sign language understanding, that performs sign language translation (SLT) and sign-subtitle alignment (SSA). Together, these two tasks enable the conversion of continuous signing videos into spoken language text and also the temporal alignment of signing with subtitles -- both essential for practical communication, large-scale corpus construction, and educational applications. To achieve this, our approach is built upon three components: (i) a lightweight visual backbone that captures manual and non-manual cues from human keypoints and lip-region images while preserving signer privacy; (ii) a Sliding Perceiver mapping network that aggregates consecutive visual features into word-level embeddings to bridge the vision-text gap; and (iii) a multi-task scalable training strategy that jointly optimises SLT and SSA, reinforcing both linguistic and temporal alignment. To promote cross-linguistic generalisation, we pretrain our model on large-scale sign-text corpora covering British Sign Language (BSL) and American Sign Language (ASL) from the BOBSL and YouTube-SL-25 datasets. With this multilingual pretraining and strong model design, we achieve state-of-the-art results on the challenging BOBSL (BSL) dataset for both SLT and SSA. Our model also demonstrates robust zero-shot generalisation and finetuned SLT performance on How2Sign (ASL), highlighting the potential of scalable translation across different sign languages. --- ### SSplain: Sparse and Smooth Explainer for Retinopathy of Prematurity Classification Authors: Elifnur Sunger, Tales Imbiriba, Peter Campbell, Deniz Erdogmus, Stratis Ioannidis, Jennifer Dy Published: 2025-12-08 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.08038v1 Neural networks are frequently used in medical diagnosis. However, due to their black-box nature, model explainers are used to help clinicians understand better and trust model outputs. This paper introduces an explainer method for classifying Retinopathy of Prematurity (ROP) from fundus images. Previous methods fail to generate explanations that preserve input image structures such as smoothness and sparsity. We introduce Sparse and Smooth Explainer (SSplain), a method that generates pixel-wise explanations while preserving image structures by enforcing smoothness and sparsity. This results in realistic explanations to enhance the understanding of the given black-box model. To achieve this goal, we define an optimization problem with combinatorial constraints and solve it using the Alternating Direction Method of Multipliers (ADMM). Experimental results show that SSplain outperforms commonly used explainers in terms of both post-hoc accuracy and smoothness analyses. Additionally, SSplain identifies features that are consistent with domain-understandable features that clinicians consider as discriminative factors for ROP. We also show SSplain's generalization by applying it to additional publicly available datasets. Code is available at https://github.com/neu-spiral/SSplain. --- ### Joint Activity Design Heuristics for Enhancing Human-Machine Collaboration Authors: Mohammadreza Jalaeian, Dane A. Morey, Michael F. Rayo Published: 2025-12-08 Categories: cs.HC, cs.AI, eess.SY arXiv: https://arxiv.org/abs/2512.08036v1 Joint activity describes when more than one agent (human or machine) contributes to the completion of a task or activity. Designing for joint activity focuses on explicitly supporting the interdependencies between agents necessary for effective coordination among agents engaged in the joint activity. This builds and expands upon designing for usability to further address how technologies can be designed to act as effective team players. Effective joint activity requires supporting, at minimum, five primary macrocognitive functions within teams: Event Detection, Sensemaking, Adaptability, Perspective-Shifting, and Coordination. Supporting these functions is equally as important as making technologies usable. We synthesized fourteen heuristics from relevant literature including display design, human factors, cognitive systems engineering, cognitive psychology, and computer science to aid the design, development, and evaluation of technologies that support joint human-machine activity. --- ### CLARITY: Medical World Model for Guiding Treatment Decisions by Simulating Context-Aware Disease Trajectories Authors: Tianxingjian Ding, Yuanhao Zou, Chen Chen, Mubarak Shah, Yu Tian Published: 2025-12-08 Categories: cs.LG, cs.CV arXiv: https://arxiv.org/abs/2512.08029v4 Clinical decision-making in oncology requires forecasting how disease evolves under treatment, yet most AI systems remain static predictors that cannot model longitudinal, treatment-conditioned progression. Although generative and world models have demonstrated strong capabilities in general domains, their adaptation to medicine remains limited and insufficient for capturing complex, treatment-induced physiological dynamics across temporal scales. To address these gaps, we introduce CLARITY, a medical world model that enables counterfactual simulation of treatment-conditioned disease trajectories for clinical decision-making. By jointly encoding imaging-derived latent states, temporal intervals that capture irregular follow-ups, and patient-specific clinical context, CLARITY learns smooth and interpretable representations of disease progression, allowing the model to simulate how alternative treatments reshape future disease dynamics. Because treatment optimization is inherently sequential and uncertain, requiring evaluation of long-term outcomes across multiple possible interventions, we further propose an entropy-regularized, computationally efficient long-horizon prediction-to-decision framework that plans treatment strategies over imagined disease trajectories and iteratively refines therapy proposals through survival-aware feedback, forming a closed-loop simulation-to-decision framework for treatment planning. CLARITY achieves state-of-the-art performance in treatment planning and survival prediction across three cancer datasets, including two brain tumor cohorts (MU-Glioma-Post and zero-shot on UCSF-ALPTDG) and one breast cancer dataset (ISPY-2), demonstrating strong generalization across cancer types while consistently outperforming prior generative methods and medical-domain large language model baselines. --- ### Toward an AI Reasoning-Enabled System for Patient-Clinical Trial Matching Authors: Caroline N. Leach, Mitchell A. Klusty, Samuel E. Armstrong, Justine C. Pickarski, Kristen L. Hankins, Emily B. Collier, Maya Shah, Aaron D. Mullen, V. K. Cody Bumgardner Published: 2025-12-08 Categories: cs.AI arXiv: https://arxiv.org/abs/2512.08026v1 Screening patients for clinical trial eligibility remains a manual, time-consuming, and resource-intensive process. We present a secure, scalable proof-of-concept system for Artificial Intelligence (AI)-augmented patient-trial matching that addresses key implementation challenges: integrating heterogeneous electronic health record (EHR) data, facilitating expert review, and maintaining rigorous security standards. Leveraging open-source, reasoning-enabled large language models (LLMs), the system moves beyond binary classification to generate structured eligibility assessments with interpretable reasoning chains that support human-in-the-loop review. This decision support tool represents eligibility as a dynamic state rather than a fixed determination, identifying matches when available and offering actionable recommendations that could render a patient eligible in the future. The system aims to reduce coordinator burden, intelligently broaden the set of trials considered for each patient and guarantee comprehensive auditability of all AI-generated outputs. --- ### Provable Diffusion Posterior Sampling for Bayesian Inversion Authors: Jinyuan Chang, Chenguang Duan, Yuling Jiao, Ruoxuan Li, Jerry Zhijian Yang, Cheng Yuan Published: 2025-12-08 Categories: stat.ML, cs.LG, math.NA, math.PR, math.ST arXiv: https://arxiv.org/abs/2512.08022v2 We propose a novel diffusion-based posterior sampling method within a plug-and-play framework. Our approach constructs a probability transport from an easy-to-sample distribution to the target posterior via a diffusion process. To initialize the sampler efficiently, we introduce a warm-start strategy for the particles. The posterior score is then approximated using a Monte Carlo estimator in which samples are generated via Langevin dynamics, avoiding the heuristic approximations prevalent in prior work. The score function driving the Langevin dynamics is learned from data, enabling the model to capture rich structural features of the underlying prior. We also establish non-asymptotic error bounds in Wasserstein-2 distance guaranteeing convergence of the proposed method even for complex, multimodal posterior distributions. We corroborate our theoretical findings with numerical experiments demonstrating the effectiveness of the method across a variety of inverse problems. ---