# Research Papers — 2025-12 · bundle page 33 of 175 > Full-text markdown of papers #1601–1650 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/33 [← Previous bundle](https://aipod.dev/papers/archive/2025-12/page/32.md) [Next bundle →](https://aipod.dev/papers/archive/2025-12/page/34.md) ### Fidelity Before Structure: Verbatim Chunks Beat Lossy Artifact Extraction in Long-Conversation LLM Memory Authors: Tao An Published: 2025-12-23 Categories: cs.AI, cs.CL, cs.IR arXiv: https://arxiv.org/abs/2601.00821v4 A growing class of conversational-memory systems compresses dialogue history into structured artifacts (extracted facts, decisions, or events) on the premise that distilled structure retrieves better than raw text. We test this premise with a controlled ablation: within one fixed retrieval--rerank--reasoning pipeline, we swap only the stored representation (LLM-extracted typed artifacts versus verbatim conversation chunks), holding the model, retriever, reranker, and judge constant. Verbatim chunks win by 15.9 points on LoCoMo (43.9% vs. 28.0%) and 22.0 points on LongMemEval-S (67.4% vs. 45.4%); a 1-hop semantic graph does not recover the gap, and six confound controls reproduce the effect. The mechanism is lossy distillation, not structure per se: accuracy tracks how much source text survives in the store, and the extracted-artifact pipeline does not beat naive RAG in overall accuracy (though chunks abstain worse; see Limitations). For the extraction designs we test, structured memory should augment verbatim text rather than replace it: adding artifacts alongside chunks preserves accuracy; substituting them forfeits the gap. Code and data: https://github.com/tao-hpu/cog-canvas --- ### Calibrating LLM Judges: Linear Probes for Fast and Reliable Uncertainty Estimation Authors: Bhaktipriya Radharapu, Eshika Saxena, Kenneth Li, Chenxi Whitehouse, Adina Williams, Nicola Cancedda Published: 2025-12-23 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2512.22245v1 As LLM-based judges become integral to industry applications, obtaining well-calibrated uncertainty estimates efficiently has become critical for production deployment. However, existing techniques, such as verbalized confidence and multi-generation methods, are often either poorly calibrated or computationally expensive. We introduce linear probes trained with a Brier score-based loss to provide calibrated uncertainty estimates from reasoning judges' hidden states, requiring no additional model training. We evaluate our approach on both objective tasks (reasoning, mathematics, factuality, coding) and subjective human preference judgments. Our results demonstrate that probes achieve superior calibration compared to existing methods with $\approx10$x computational savings, generalize robustly to unseen evaluation domains, and deliver higher accuracy on high-confidence predictions. However, probes produce conservative estimates that underperform on easier datasets but may benefit safety-critical deployments prioritizing low false-positive rates. Overall, our work demonstrates that interpretability-based uncertainty estimation provides a practical and scalable plug-and-play solution for LLM judges in production. --- ### Predicting Mycotoxin Contamination in Irish Oats Using Deep and Transfer Learning Authors: Alan Inglis, Fiona Doohan, Subramani Natarajan, Breige McNulty, Chris Elliott, Anne Nugent, Julie Meneely, Brett Greer, Stephen Kildea, Diana Bucur, Martin Danaher, Melissa Di Rocco et al. Published: 2025-12-23 Categories: cs.LG, stat.ML arXiv: https://arxiv.org/abs/2512.22243v1 Mycotoxin contamination poses a significant risk to cereal crop quality, food safety, and agricultural productivity. Accurate prediction of mycotoxin levels can support early intervention strategies and reduce economic losses. This study investigates the use of neural networks and transfer learning models to predict mycotoxin contamination in Irish oat crops as a multi-response prediction task. Our dataset comprises oat samples collected in Ireland, containing a mix of environmental, agronomic, and geographical predictors. Five modelling approaches were evaluated: a baseline multilayer perceptron (MLP), an MLP with pre-training, and three transfer learning models; TabPFN, TabNet, and FT-Transformer. Model performance was evaluated using regression (RMSE, $R^2$) and classification (AUC, F1) metrics, with results reported per toxin and on average. Additionally, permutation-based variable importance analysis was conducted to identify the most influential predictors across both prediction tasks. The transfer learning approach TabPFN provided the overall best performance, followed by the baseline MLP. Our variable importance analysis revealed that weather history patterns in the 90-day pre-harvest period were the most important predictors, alongside seed moisture content. --- ### Fairness Evaluation of Risk Estimation Models for Lung Cancer Screening Authors: Shaurya Gaur, Michel Vitale, Alessa Hering, Johan Kwisthout, Colin Jacobs, Lena Philipp, Fennie van der Graaf Published: 2025-12-23 Categories: cs.LG, cs.AI, cs.CV, cs.CY, eess.IV arXiv: https://arxiv.org/abs/2512.22242v1 Lung cancer is the leading cause of cancer-related mortality in adults worldwide. Screening high-risk individuals with annual low-dose CT (LDCT) can support earlier detection and reduce deaths, but widespread implementation may strain the already limited radiology workforce. AI models have shown potential in estimating lung cancer risk from LDCT scans. However, high-risk populations for lung cancer are diverse, and these models' performance across demographic groups remains an open question. In this study, we drew on the considerations on confounding factors and ethically significant biases outlined in the JustEFAB framework to evaluate potential performance disparities and fairness in two deep learning risk estimation models for lung cancer screening: the Sybil lung cancer risk model and the Venkadesh21 nodule risk estimator. We also examined disparities in the PanCan2b logistic regression model recommended in the British Thoracic Society nodule management guideline. Both deep learning models were trained on data from the US-based National Lung Screening Trial (NLST), and assessed on a held-out NLST validation set. We evaluated AUROC, sensitivity, and specificity across demographic subgroups, and explored potential confounding from clinical risk factors. We observed a statistically significant AUROC difference in Sybil's performance between women (0.88, 95% CI: 0.86, 0.90) and men (0.81, 95% CI: 0.78, 0.84, p < .001). At 90% specificity, Venkadesh21 showed lower sensitivity for Black (0.39, 95% CI: 0.23, 0.59) than White participants (0.69, 95% CI: 0.65, 0.73). These differences were not explained by available clinical confounders and thus may be classified as unfair biases according to JustEFAB. Our findings highlight the importance of improving and monitoring model performance across underrepresented subgroups, and further research on algorithmic fairness, in lung cancer screening. --- ### Enhanced geometry prediction in laser directed energy deposition using meta-learning Authors: Abdul Malik Al Mardhouf Al Saadi, Amrita Basak Published: 2025-12-23 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2512.22241v1 Accurate bead geometry prediction in laser-directed energy deposition (L-DED) is often hindered by the scarcity and heterogeneity of experimental datasets collected under different materials, machine configurations, and process parameters. To address this challenge, a cross-dataset knowledge transfer model based on meta-learning for predicting deposited track geometry in L-DED is proposed. Specifically, two gradient-based meta-learning algorithms, i.e., Model-Agnostic Meta-Learning (MAML) and Reptile, are investigated to enable rapid adaptation to new deposition conditions with limited data. The proposed framework is performed using multiple experimental datasets compiled from peer-reviewed literature and in-house experiments and evaluated across powder-fed, wire-fed, and hybrid wire-powder L-DED processes. Results show that both MAML and Reptile achieve accurate bead height predictions on unseen target tasks using as few as three to nine training examples, consistently outperforming conventional feedforward neural networks trained under comparable data constraints. Across multiple target tasks representing different printing conditions, the meta-learning models achieve strong generalization performance, with R-squared values reaching up to approximately 0.9 and mean absolute errors between 0.03-0.08 mm, demonstrating effective knowledge transfer across heterogeneous L-DED settings. --- ### EvoXplain: When Machine Learning Models Agree on Predictions but Disagree on Why -- Measuring Mechanistic Multiplicity Across Training Runs Authors: Chama Bensmail Published: 2025-12-23 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2512.22240v5 Machine learning models are primarily judged by predictive performance, especially in applied genomics, where explanations are read as biological findings. In practice, reported gene panels are stabilised by averaging, ranking, or taking consensus over the many models a pipeline produces across cross-validation folds, tuning grids, and repeated runs. This raises an overlooked question: when two models achieve high accuracy, do they rely on the same internal logic, or reach the same outcome via different mechanisms? We introduce EvoXplain, a diagnostic framework that measures whether a pipeline's explanation is uniquely determined across repeated training and model selection. Rather than analysing a single trained model, EvoXplain treats explanations as samples drawn from the training and model selection pipeline itself, without aggregating predictions or constructing ensembles, and examines whether they form a single coherent explanatory basin or separate into multiple structured basins. We evaluate EvoXplain on a TCGA pan-cancer cohort and a within-cancer breast-cancer subtype task, using elastic-net Logistic Regression and gradient-boosted trees. Although all models reach about 98% accuracy, explanation structure differs across pipelines. Holding the data split fixed and varying only the regularisation strength, equally accurate Logistic Regression models separate into a few discrete, reproducible basins that recur across 100 data splits and carry distinct biological content, while the gradient-boosted pipeline converges to one basin. The same multiplicity appears within a single cancer subtype, from the ordinary tuning step alone. EvoXplain makes explanatory structure visible, revealing when an averaged consensus corresponds to no single trained model, and reframes interpretability as a property of the training pipeline rather than of any single model. --- ### Multi-objective hybrid knowledge distillation for efficient deep learning in smart agriculture Authors: Phi-Hung Hoang, Nam-Thuan Trinh, Van-Manh Tran, Thi-Thu-Hong Phan Published: 2025-12-23 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2512.22239v1 Deploying deep learning models on resource-constrained edge devices remains a major challenge in smart agriculture due to the trade-off between computational efficiency and recognition accuracy. To address this challenge, this study proposes a hybrid knowledge distillation framework for developing a lightweight yet high-performance convolutional neural network. The proposed approach designs a customized student model that combines inverted residual blocks with dense connectivity and trains it under the guidance of a ResNet18 teacher network using a multi-objective strategy that integrates hard-label supervision, feature-level distillation, response-level distillation, and self-distillation. Experiments are conducted on a rice seed variety identification dataset containing nine varieties and further extended to four plant leaf disease datasets, including rice, potato, coffee, and corn, to evaluate generalization capability. On the rice seed variety classification task, the distilled student model achieves an accuracy of 98.56%, which is only 0.09% lower than the teacher model (98.65%), while requiring only 0.68 GFLOPs and approximately 1.07 million parameters. This corresponds to a reduction of about 2.7 times in computational cost and more than 10 times in model size compared with the ResNet18 teacher model. In addition, compared with representative pretrained models, the proposed student reduces the number of parameters by more than 6 times relative to DenseNet121 and by over 80 times compared with the Vision Transformer (ViT) architecture, while maintaining comparable or superior classification accuracy. Consistent performance gains across multiple plant leaf disease datasets further demonstrate the robustness, efficiency, and strong deployment potential of the proposed framework for hardware-limited smart agriculture systems. --- ### Masking Teacher and Reinforcing Student for Distilling Vision-Language Models Authors: Byung-Kwan Lee, Yu-Chiang Frank Wang, Ryo Hachiuma Published: 2025-12-23 Categories: cs.LG, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2512.22238v1 Large-scale vision-language models (VLMs) have recently achieved remarkable multimodal understanding, but their massive size makes them impractical for deployment on mobile or edge devices. This raises the need for compact yet capable VLMs that can efficiently learn from powerful large teachers. However, distilling knowledge from a large teacher to a small student remains challenging due to their large size gap: the student often fails to reproduce the teacher's complex, high-dimensional representations, leading to unstable learning and degraded performance. To address this, we propose Masters (Masking Teacher and Reinforcing Student), a mask-progressive reinforcement learning (RL) distillation framework. Masters first masks non-dominant weights of the teacher to reduce unnecessary complexity, then progressively restores the teacher by gradually increasing its capacity during training. This strategy allows the student to learn richer representations from the teacher in a smooth and stable manner. To further refine knowledge transfer, Masters integrates an offline RL stage with two complementary rewards: an accuracy reward that measures the correctness of the generated responses, and a distillation reward that quantifies the ease of transferring responses from teacher to student. Unlike online think-answer RL paradigms that are computationally expensive and generate lengthy responses, our offline RL leverages pre-generated responses from masked teachers. These provide rich yet efficient guidance, enabling students to achieve strong performance without requiring the think-answer process. --- ### Meta-information Guided Cross-domain Synergistic Diffusion Model for Low-dose PET Reconstruction Authors: Mengxiao Geng, Ran Hong, Xiaoling Xu, Bingxuan Li, Qiegen Liu Published: 2025-12-23 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2512.22237v1 Low-dose PET imaging is crucial for reducing patient radiation exposure but faces challenges like noise interference, reduced contrast, and difficulty in preserving physiological details. Existing methods often neglect both projection-domain physics knowledge and patient-specific meta-information, which are critical for functional-semantic correlation mining. In this study, we introduce a meta-information guided cross-domain synergistic diffusion model (MiG-DM) that integrates comprehensive cross-modal priors to generate high-quality PET images. Specifically, a meta-information encoding module transforms clinical parameters into semantic prompts by considering patient characteristics, dose-related information, and semi-quantitative parameters, enabling cross-modal alignment between textual meta-information and image reconstruction. Additionally, the cross-domain architecture combines projection-domain and image-domain processing. In the projection domain, a specialized sinogram adapter captures global physical structures through convolution operations equivalent to global image-domain filtering. Experiments on the UDPET public dataset and clinical datasets with varying dose levels demonstrate that MiG-DM outperforms state-of-the-art methods in enhancing PET image quality and preserving physiological details. --- ### We are not able to identify AI-generated images Authors: Adrien Pavão Published: 2025-12-23 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2512.22236v1 AI-generated images are now pervasive online, yet many people believe they can easily tell them apart from real photographs. We test this assumption through an interactive web experiment where participants classify 20 images as real or AI-generated. Our dataset contains 120 difficult cases: real images sampled from CC12M, and carefully curated AI-generated counterparts produced with MidJourney. In total, 165 users completed 233 sessions. Their average accuracy was 54%, only slightly above random guessing, with limited improvement across repeated attempts. Response times averaged 7.3 seconds, and some images were consistently more deceptive than others. These results indicate that, even on relatively simple portrait images, humans struggle to reliably detect AI-generated content. As synthetic media continues to improve, human judgment alone is becoming insufficient for distinguishing real from artificial data. These findings highlight the need for greater awareness and ethical guidelines as AI-generated media becomes increasingly indistinguishable from reality. --- ### DiRL: An Efficient Post-Training Framework for Diffusion Language Models Authors: Ying Zhu, Jiaxin Wan, Xiaoran Liu, Siyang He, Qiqi Wang, Xu Guo, Tianyi Liang, Zengfeng Huang, Ziwei He, Xipeng Qiu Published: 2025-12-23 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2512.22234v2 Diffusion Language Models (dLLMs) have emerged as promising alternatives to Auto-Regressive (AR) models. While recent efforts have validated their pre-training potential and accelerated inference speeds, the post-training landscape for dLLMs remains underdeveloped. Existing methods suffer from computational inefficiency and objective mismatches between training and inference, severely limiting performance on complex reasoning tasks such as mathematics. To address this, we introduce DiRL, an efficient post-training framework that tightly integrates FlexAttention-accelerated blockwise training with LMDeploy-optimized inference. This architecture enables a streamlined online model update loop, facilitating efficient two-stage post-training (Supervised Fine-Tuning followed by Reinforcement Learning). Building on this framework, we propose DiPO, the first unbiased Group Relative Policy Optimization (GRPO) implementation tailored for dLLMs. We validate our approach by training DiRL-8B-Instruct on high-quality math data. Our model achieves state-of-the-art math performance among dLLMs and surpasses comparable models in the Qwen2.5 series on several benchmarks. --- ### Scalable Cloud-Native Architectures for Intelligent PMU Data Processing Authors: Nachiappan Chockalingam, Akshay Deshpande, Lokesh Butra, Ram Sekhar Bodala, Nitin Saksena, Adithya Parthasarathy, Balakrishna Pothineni, Akash Kumar Agarwal Published: 2025-12-23 Categories: cs.DC, cs.AI arXiv: https://arxiv.org/abs/2512.22231v1 Phasor Measurement Units (PMUs) generate high-frequency, time-synchronized data essential for real-time power grid monitoring, yet the growing scale of PMU deployments creates significant challenges in latency, scalability, and reliability. Conventional centralized processing architectures are increasingly unable to handle the volume and velocity of PMU data, particularly in modern grids with dynamic operating conditions. This paper presents a scalable cloud-native architecture for intelligent PMU data processing that integrates artificial intelligence with edge and cloud computing. The proposed framework employs distributed stream processing, containerized microservices, and elastic resource orchestration to enable low-latency ingestion, real-time anomaly detection, and advanced analytics. Machine learning models for time-series analysis are incorporated to enhance grid observability and predictive capabilities. Analytical models are developed to evaluate system latency, throughput, and reliability, showing that the architecture can achieve sub-second response times while scaling to large PMU deployments. Security and privacy mechanisms are embedded to support deployment in critical infrastructure environments. The proposed approach provides a robust and flexible foundation for next-generation smart grid analytics. --- ### KAN-FPN-Stem:A KAN-Enhanced Feature Pyramid Stem for Boosting ViT-based Pose Estimation Authors: HaoNan Tang Published: 2025-12-23 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.22228v1 Vision Transformers (ViT) have demonstrated significant promise in dense prediction tasks such as pose estimation. However, their performance is frequently constrained by the overly simplistic front-end designs employed in models like ViTPose. This naive patchification mechanism struggles to effectively handle multi-scale variations and results in irreversible information loss during the initial feature extraction phase. To overcome this limitation, we introduce a novel KAN-enhanced FPN-Stem architecture. Through rigorous ablation studies, we first identified that the true bottleneck for performance improvement lies not in plug-and-play attention modules (e.g., CBAM), but in the post-fusion non-linear smoothing step within the FPN. Guided by this insight, our core innovation is to retain the classic "upsample-and-add" fusion stream of the FPN, but replace its terminal, standard linear 3x3 smoothing convolution with a powerful KAN-based convolutional layer. Leveraging its superior non-linear modeling capabilities, this KAN-based layer adaptively learns and rectifies the "artifacts" generated during the multi-scale fusion process. Extensive experiments on the COCO dataset demonstrate that our KAN-FPN-Stem achieves a significant performance boost of up to +2.0 AP over the lightweight ViTPose-S baseline. This work not only delivers a plug-and-play, high-performance module but, more importantly, reveals that: the performance bottleneck in ViT front-end often lies not in 'feature refinement' (Attention), but in the quality of 'feature fusion' (Fusion). Furthermore, it provides an effective path to address this bottleneck through the introduction of the KAN operator. --- ### Probing Spectrum-Like Organization of States of Mind in Transformer Representation Spaces Authors: Sophie Zhao Published: 2025-12-23 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2512.22227v3 We investigate whether graded states of mind form spectrum-like structure in transformer representation spaces. To do so, we construct a dataset of 636 short natural-language sentences annotated with both a continuous score from $-5$ to $5$ and one of seven ordered tiers, ranging from collapsed or scarcity-driven expressions to more coherent, reflective, and integrative ones. We evaluate five frozen transformer representations: four sentence-embedding models and one decoder-only residual-stream representation. Across all representations, simple probes reliably recover both the continuous score and the discrete tier labels, and permutation tests show that performance significantly exceeds shuffled-label baselines. Additional analyses reveal a consistent geometric pattern: UMAP projections show low-to-high organization, confusion matrices concentrate errors between neighboring tiers, and directional ablation identifies a prominent score-aligned component. These results suggest that transformer representations contain statistically significant, spectrum-like organization aligned with the annotated state-of-mind structure. The annotations are used only as an operational framework for representation analysis, not as a clinical or diagnostic measure. --- ### VideoScaffold: Elastic-Scale Visual Hierarchies for Streaming Video Understanding in MLLMs Authors: Naishan Zheng, Jie Huang, Qingpei Guo, Feng Zhao Published: 2025-12-23 Categories: cs.CV, cs.AI, cs.CL, cs.LG arXiv: https://arxiv.org/abs/2512.22226v1 Understanding long videos with multimodal large language models (MLLMs) remains challenging due to the heavy redundancy across frames and the need for temporally coherent representations. Existing static strategies, such as sparse sampling, frame compression, and clustering, are optimized for offline settings and often produce fragmented or over-compressed outputs when applied to continuous video streams. We present VideoScaffold, a dynamic representation framework designed for streaming video understanding. It adaptively adjusts event granularity according to video duration while preserving fine-grained visual semantics. VideoScaffold introduces two key components: Elastic-Scale Event Segmentation (EES), which performs prediction-guided segmentation to dynamically refine event boundaries, and Hierarchical Event Consolidation (HEC), which progressively aggregates semantically related segments into multi-level abstractions. Working in concert, EES and HEC enable VideoScaffold to transition smoothly from fine-grained frame understanding to abstract event reasoning as the video stream unfolds. Extensive experiments across both offline and streaming video understanding benchmarks demonstrate that VideoScaffold achieves state-of-the-art performance. The framework is modular and plug-and-play, seamlessly extending existing image-based MLLMs to continuous video comprehension. The code is available at https://github.com/zheng980629/VideoScaffold. --- ### Literature Mining System for Nutraceutical Biosynthesis: From AI Framework to Biological Insight Authors: Xinyang Sun, Nipon Sarmah, Miao Guo Published: 2025-12-23 Categories: q-bio.QM, cs.AI arXiv: https://arxiv.org/abs/2512.22225v1 The extraction of structured knowledge from scientific literature remains a major bottleneck in nutraceutical research, particularly when identifying microbial strains involved in compound biosynthesis. This study presents a domain-adapted system powered by large language models (LLMs) and guided by advanced prompt engineering techniques to automate the identification of nutraceutical-producing microbes from unstructured scientific text. By leveraging few-shot prompting and tailored query designs, the system demonstrates robust performance across multiple configurations, with DeepSeekV3 outperforming LLaMA2 in accuracy, especially when domain-specific strain information is included. A structured and validated dataset comprising 35 nutraceutical-strain associations was generated, spanning amino acids, fibers, phytochemicals, and vitamins. The results reveal significant microbial diversity across monoculture and co-culture systems, with dominant contributions from Corynebacterium glutamicum, Escherichia coli, and Bacillus subtilis, alongside emerging synthetic consortia. This AI-driven framework not only enhances the scalability and interpretability of literature mining but also provides actionable insights for microbial strain selection, synthetic biology design, and precision fermentation strategies in the production of high-value nutraceuticals. --- ### ReGAIN: Retrieval-Grounded AI Framework for Network Traffic Analysis Authors: Shaghayegh Shajarian, Kennedy Marsh, James Benson, Sajad Khorsandroo, Mahmoud Abdelsalam Published: 2025-12-23 Categories: cs.LG, cs.AI, cs.CR, cs.NI arXiv: https://arxiv.org/abs/2512.22223v1 Modern networks generate vast, heterogeneous traffic that must be continuously analyzed for security and performance. Traditional network traffic analysis systems, whether rule-based or machine learning-driven, often suffer from high false positives and lack interpretability, limiting analyst trust. In this paper, we present ReGAIN, a multi-stage framework that combines traffic summarization, retrieval-augmented generation (RAG), and Large Language Model (LLM) reasoning for transparent and accurate network traffic analysis. ReGAIN creates natural-language summaries from network traffic, embeds them into a multi-collection vector database, and utilizes a hierarchical retrieval pipeline to ground LLM responses with evidence citations. The pipeline features metadata-based filtering, MMR sampling, a two-stage cross-encoder reranking mechanism, and an abstention mechanism to reduce hallucinations and ensure grounded reasoning. Evaluated on ICMP ping flood and TCP SYN flood traces from the real-world traffic dataset, it demonstrates robust performance, achieving accuracy between 95.95% and 98.82% across different attack types and evaluation benchmarks. These results are validated against two complementary sources: dataset ground truth and human expert assessments. ReGAIN also outperforms rule-based, classical ML, and deep learning baselines while providing unique explainability through trustworthy, verifiable responses. --- ### A Study of Solving Life-and-Death Problems in Go Using Relevance-Zone Based Solvers Authors: Chung-Chin Shih, Ti-Rong Wu, Ting Han Wei, Yu-Shan Hsu, Hung Guei, I-Chen Wu Published: 2025-12-23 Categories: cs.AI arXiv: https://arxiv.org/abs/2512.21365v1 This paper analyzes the behavior of solving Life-and-Death (L&D) problems in the game of Go using current state-of-the-art computer Go solvers with two techniques: the Relevance-Zone Based Search (RZS) and the relevance-zone pattern table. We examined the solutions derived by relevance-zone based solvers on seven L&D problems from the renowned book "Life and Death Dictionary" written by Cho Chikun, a Go grandmaster, and found several interesting results. First, for each problem, the solvers identify a relevance-zone that highlights the critical areas for solving. Second, the solvers discover a series of patterns, including some that are rare. Finally, the solvers even find different answers compared to the given solutions for two problems. We also identified two issues with the solver: (a) it misjudges values of rare patterns, and (b) it tends to prioritize living directly rather than maximizing territory, which differs from the behavior of human Go players. We suggest possible approaches to address these issues in future work. Our code and data are available at https://rlg.iis.sinica.edu.tw/papers/study-LD-RZ. --- ### From Visual Perception to Deep Empathy: An Automated Assessment Framework for House-Tree-Person Drawings Using Multimodal LLMs and Multi-Agent Collaboration Authors: Shuide Wen, Yu Sun, Beier Ku, Zhi Gao, Lijun Ma, Yang Yang, Can Jiao Published: 2025-12-23 Categories: cs.AI, eess.SY arXiv: https://arxiv.org/abs/2512.21360v1 Background: The House-Tree-Person (HTP) drawing test, introduced by John Buck in 1948, remains a widely used projective technique in clinical psychology. However, it has long faced challenges such as heterogeneous scoring standards, reliance on examiners subjective experience, and a lack of a unified quantitative coding system. Results: Quantitative experiments showed that the mean semantic similarity between Multimodal Large Language Model (MLLM) interpretations and human expert interpretations was approximately 0.75 (standard deviation about 0.05). In structurally oriented expert data sets, this similarity rose to 0.85, indicating expert-level baseline comprehension. Qualitative analyses demonstrated that the multi-agent system, by integrating social-psychological perspectives and destigmatizing narratives, effectively corrected visual hallucinations and produced psychological reports with high ecological validity and internal coherence. Conclusions: The findings confirm the potential of multimodal large models as standardized tools for projective assessment. The proposed multi-agent framework, by dividing roles, decouples feature recognition from psychological inference and offers a new paradigm for digital mental-health services. Keywords: House-Tree-Person test; multimodal large language model; multi-agent collaboration; cosine similarity; computational psychology; artificial intelligence --- ### Nemotron 3 Nano: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning Authors: NVIDIA, :, Aaron Blakeman, Aaron Grattafiori, Aarti Basant, Abhibha Gupta, Abhinav Khattar, Adi Renduchintala, Aditya Vavre, Akanksha Shukla, Akhiad Bercovich, Aleksander Ficek et al. Published: 2025-12-23 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2512.20848v1 We present Nemotron 3 Nano 30B-A3B, a Mixture-of-Experts hybrid Mamba-Transformer language model. Nemotron 3 Nano was pretrained on 25 trillion text tokens, including more than 3 trillion new unique tokens over Nemotron 2, followed by supervised fine tuning and large-scale RL on diverse environments. Nemotron 3 Nano achieves better accuracy than our previous generation Nemotron 2 Nano while activating less than half of the parameters per forward pass. It achieves up to 3.3x higher inference throughput than similarly-sized open models like GPT-OSS-20B and Qwen3-30B-A3B-Thinking-2507, while also being more accurate on popular benchmarks. Nemotron 3 Nano demonstrates enhanced agentic, reasoning, and chat abilities and supports context lengths up to 1M tokens. We release both our pretrained Nemotron 3 Nano 30B-A3B Base and post-trained Nemotron 3 Nano 30B-A3B checkpoints on Hugging Face. --- ### MAR:Multi-Agent Reflexion Improves Reasoning Abilities in LLMs Authors: Onat Ozer, Yuchen Wang, Grace Wu, Daniel Dosti, Honghao Zhang, Vivi De La Rue Published: 2025-12-23 Categories: cs.AI, cs.MA arXiv: https://arxiv.org/abs/2512.20845v2 LLMs have shown the capacity to improve their performance on reasoning tasks through reflecting on their mistakes, and acting with these reflections in mind. However, continual reflections of the same LLM onto itself exhibit degeneration of thought, where the LLM continues to repeat the same errors again and again even with the knowledge that its wrong. To address this problem, we instead introduce multi-agent with multi-persona debators as the method to generate reflections. Through out extensive experimentation, we've found that the leads to better diversity of in the reflections generated by the llm agent. We demonstrate an accuracy of 47% EM HotPot QA (question answering) and 82.7% on HumanEval (programming), both performances surpassing reflection with a single llm. --- ### Input-Adaptive Visual Preprocessing for Efficient Fast Vision-Language Model Inference Authors: Putu Indah Githa Cahyani, Komang David Dananjaya Suartana, Novanto Yudistira Published: 2025-12-23 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.20839v1 Vision-Language Models (VLMs) have demonstrated strong performance on multimodal reasoning tasks, but their deployment remains challenging due to high inference latency and computational cost, particularly when processing high-resolution visual inputs. While recent architectures such as FastVLM improve efficiency through optimized vision encoders, existing pipelines still rely on static visual preprocessing, leading to redundant computation for visually simple inputs. In this work, we propose an adaptive visual preprocessing method that dynamically adjusts input resolution and spatial coverage based on image content characteristics. The proposed approach combines content-aware image analysis, adaptive resolution selection, and content-aware cropping to reduce visual redundancy prior to vision encoding. Importantly, the method is integrated with FastVLM without modifying its architecture or requiring retraining. We evaluate the proposed method on a subset of the DocVQA dataset in an inference-only setting, focusing on efficiency-oriented metrics. Experimental results show that adaptive preprocessing reduces per-image inference time by over 50\%, lowers mean full generation time, and achieves a consistent reduction of more than 55\% in visual token count compared to the baseline pipeline. These findings demonstrate that input-aware preprocessing is an effective and lightweight strategy for improving deployment-oriented efficiency of vision-language models. To facilitate reproducibility, our implementation is provided as a fork of the FastVLM repository, incorporating the files for the proposed method, and is available at https://github.com/kmdavidds/mlfastlm. --- ### CHAMMI-75: Pre-training multi-channel models with heterogeneous microscopy images Authors: Vidit Agrawal, John Peters, Tyler N. Thompson, Mohammad Vali Sanian, Chau Pham, Nikita Moshkov, Arshad Kazi, Aditya Pillai, Jack Freeman, Byunguk Kang, Samouil L. Farhi, Ernest Fraenkel et al. Published: 2025-12-23 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2512.20833v2 Quantifying cell morphology using images and machine learning has proven to be a powerful tool to study the response of cells to treatments. However, models used to quantify cellular morphology are typically trained with a single microscopy imaging type. This results in specialized models that cannot be reused across biological studies because the technical specifications do not match (e.g., different number of channels). Here, we present CHAMMI-75, an open access dataset of heterogeneous, multi-channel microscopy images from 75 diverse biological studies. We curated this resource from publicly available sources to investigate cellular morphology models that are channel-adaptive and can process any microscopy image type. Our experiments show that training with CHAMMI-75 can improve performance in multi-channel bioimaging tasks primarily because of its high diversity in microscopy modalities. This work paves the way to create the next generation of cellular morphology models for biological studies. --- ### Context-Sensitive Abstractions for Reinforcement Learning with Parameterized Actions Authors: Rashmeet Kaur Nayyar, Naman Shah, Siddharth Srivastava Published: 2025-12-23 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2512.20831v2 Real-world sequential decision-making often involves parameterized action spaces that require both, decisions regarding discrete actions and decisions about continuous action parameters governing how an action is executed. Existing approaches exhibit severe limitations in this setting -- planning methods demand hand-crafted action models, and standard reinforcement learning (RL) algorithms are designed for either discrete or continuous actions but not both, and the few RL methods that handle parameterized actions typically rely on domain-specific engineering and fail to exploit the latent structure of these spaces. This paper extends the scope of RL algorithms to long-horizon, sparse-reward settings with parameterized actions by enabling agents to autonomously learn both state and action abstractions online. We introduce algorithms that progressively refine these abstractions during learning, increasing fine-grained detail in the critical regions of the state-action space where greater resolution improves performance. Across several continuous-state, parameterized-action domains, our abstraction-driven approach enables TD($λ$) to achieve markedly higher sample efficiency than state-of-the-art baselines. --- ### NotSoTiny: A Large, Living Benchmark for RTL Code Generation Authors: Razine Moundir Ghorab, Emanuele Parisi, Cristian Gutierrez, Miquel Alberti-Binimelis, Miquel Moreto, Dario Garcia-Gasulla, Gokcen Kestor Published: 2025-12-23 Categories: cs.AR, cs.AI arXiv: https://arxiv.org/abs/2512.20823v1 LLMs have shown early promise in generating RTL code, yet evaluating their capabilities in realistic setups remains a challenge. So far, RTL benchmarks have been limited in scale, skewed toward trivial designs, offering minimal verification rigor, and remaining vulnerable to data contamination. To overcome these limitations and to push the field forward, this paper introduces NotSoTiny, a benchmark that assesses LLM on the generation of structurally rich and context-aware RTL. Built from hundreds of actual hardware designs produced by the Tiny Tapeout community, our automated pipeline removes duplicates, verifies correctness and periodically incorporates new designs to mitigate contamination, matching Tiny Tapeout release schedule. Evaluation results show that NotSoTiny tasks are more challenging than prior benchmarks, emphasizing its effectiveness in overcoming current limitations of LLMs applied to hardware design, and in guiding the improvement of such promising technology. --- ### MediEval: A Unified Medical Benchmark for Patient-Contextual and Knowledge-Grounded Reasoning in LLMs Authors: Zhan Qu, Michael Färber Published: 2025-12-23 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2512.20822v2 Large Language Models (LLMs) are increasingly applied to medicine, yet their adoption is limited by concerns over reliability and safety. Existing evaluations either test factual medical knowledge in isolation or assess patient-level reasoning without verifying correctness, leaving a critical gap. We introduce MediEval, a benchmark that links MIMIC-IV electronic health records (EHRs) to a unified knowledge base built from UMLS and other biomedical vocabularies. MediEval generates diverse factual and counterfactual medical statements within real patient contexts, enabling systematic evaluation across a 4-quadrant framework that jointly considers knowledge grounding and contextual consistency. Using this framework, we identify critical failure modes, including hallucinated support and truth inversion, that current proprietary, open-source, and domain-specific LLMs frequently exhibit. To address these risks, we propose Counterfactual Risk-Aware Fine-tuning (CoRFu), a DPO-based method with an asymmetric penalty targeting unsafe confusions. CoRFu improves by +16.4 macro-F1 points over the base model and eliminates truth inversion errors, demonstrating both higher accuracy and substantially greater safety. --- ### Divided We Fall: Defending Against Adversarial Attacks via Soft-Gated Fractional Mixture-of-Experts with Randomized Adversarial Training Authors: Mohammad Meymani, Roozbeh Razavi-Far Published: 2025-12-23 Categories: cs.LG arXiv: https://arxiv.org/abs/2512.20821v3 Machine learning is a powerful tool enabling full automation of a huge number of tasks without explicit programming. Despite recent progress of machine learning in different domains, these models have shown vulnerabilities when they are exposed to adversarial threats. Adversarial threats aim to hinder the machine learning models from satisfying their objectives. They can create adversarial perturbations, which are imperceptible to humans' eyes but have the ability to cause misclassification during inference. In this paper, we propose a defense system, which devises an adversarial training module within mixture-of-experts architecture to enhance its robustness against white-box evasion attacks. In our proposed defense system, we use nine pre-trained classifiers (experts) with ResNet-18 as their backbone. During end-to-end training, the parameters of all experts and the gating mechanism are jointly updated allowing further optimization of the experts. Our proposed defense system outperforms prior MoE-based defenses under strong white-box FGSM and PGD evaluation on CIFAR-10 and SVHN. The use of multiple experts increases training time and compute relative to single-network baselines; however, inference scales approximately linearly with the number of experts and is substantially cheaper than training. --- ### EssayCBM: Rubric-Aligned Concept Bottleneck Models for Transparent Essay Grading Authors: Kumar Satvik Chaudhary, Chengshuai Zhao, Fan Zhang, Garima Agrawal, Yuli Deng, Huan Liu Published: 2025-12-23 Categories: cs.CL arXiv: https://arxiv.org/abs/2512.20817v3 Automated essay scoring (AES) has advanced significantly with neural language models, yet most systems remain opaque, offering little visibility into how grades are produced. In educational settings, instructors must be able to understand, trust, and occasionally override the automated grading decisions. We introduce EssayCBM, a rubric-aligned concept bottleneck framework that decomposes essay evaluation into eight interpretable writing concepts before computing the final score. Unlike direct LLM-based grading approaches, EssayCBM learns an explicit and auditable mapping from writing concepts to grades, allowing instructors to inspect and adjust rubric-level predictions during grading. EssayCBM matches neural AES baselines while making grading decisions transparent and directly editable at the rubric level. We further present an interactive system that demonstrates this capability by allowing instructors to inspect and modify concept predictions in real time. --- ### Learning to Sense for Driving: Joint Optics-Sensor-Model Co-Design for Semantic Segmentation Authors: Reeshad Khan, John Gauch Published: 2025-12-23 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.20815v2 Traditional autonomous driving pipelines decouple camera design from downstream perception, relying on fixed optics and handcrafted ISPs that prioritize human viewable imagery rather than machine semantics. This separation discards information during demosaicing, denoising, or quantization, while forcing models to adapt to sensor artifacts. We present a task-driven co-design framework that unifies optics, sensor modeling, and lightweight semantic segmentation networks into a single end-to-end RAW-to-task pipeline. Building on DeepLens(19), our system integrates realistic cellphone-scale lens models, learnable color filter arrays, Poisson-Gaussian noise processes, and quantization, all optimized directly for segmentation objectives. Evaluations on KITTI-360 show consistent mIoU improvements over fixed pipelines, with optics modeling and CFA learning providing the largest gains, especially for thin or low-light-sensitive classes. Importantly, these robustness gains are achieved with a compact ~1M-parameter model running at ~28 FPS, demonstrating edge deployability. Visual and quantitative analyses further highlight how co-designed sensors adapt acquisition to semantic structure, sharpening boundaries and maintaining accuracy under blur, noise, and low bit-depth. Together, these findings establish full-stack co-optimization of optics, sensors, and networks as a principled path toward efficient, reliable, and deployable perception in autonomous systems. --- ### FedMPDD: Communication-Efficient Federated Learning with Privacy Preservation Attributes via Projected Directional Derivative Authors: Mohammadreza Rostami, Solmaz S. Kia Published: 2025-12-23 Categories: cs.LG arXiv: https://arxiv.org/abs/2512.20814v1 This paper introduces \texttt{FedMPDD} (\textbf{Fed}erated Learning via \textbf{M}ulti-\textbf{P}rojected \textbf{D}irectional \textbf{D}erivatives), a novel algorithm that simultaneously optimizes bandwidth utilization and enhances privacy in Federated Learning. The core idea of \texttt{FedMPDD} is to encode each client's high-dimensional gradient by computing its directional derivatives along multiple random vectors. This compresses the gradient into a much smaller message, significantly reducing uplink communication costs from $\mathcal{O}(d)$ to $\mathcal{O}(m)$, where $m \ll d$. The server then decodes the aggregated information by projecting it back onto the same random vectors. Our key insight is that averaging multiple projections overcomes the dimension-dependent convergence limitations of a single projection. We provide a rigorous theoretical analysis, establishing that \texttt{FedMPDD} converges at a rate of $\mathcal{O}(1/\sqrt{K})$, matching the performance of FedSGD. Furthermore, we demonstrate that our method provides some inherent privacy against gradient inversion attacks due to the geometric properties of low-rank projections, offering a tunable privacy-utility trade-off controlled by the number of projections. Extensive experiments on benchmark datasets validate our theory and demonstrates our results. --- ### GraphFire-X: Physics-Informed Graph Attention Networks and Structural Gradient Boosting for Building-Scale Wildfire Preparedness at the Wildland-Urban Interface Authors: Miguel Esparza, Vamshi Battal, Ali Mostafavi Published: 2025-12-23 Categories: cs.LG arXiv: https://arxiv.org/abs/2512.20813v1 As wildfires increasingly evolve into urban conflagrations, traditional risk models that treat structures as isolated assets fail to capture the non-linear contagion dynamics characteristic of the wildland urban interface (WUI). This research bridges the gap between mechanistic physics and data driven learning by establishing a novel dual specialist ensemble framework that disentangles vulnerability into two distinct vectors, environmental contagion and structural fragility. The architecture integrates two specialized predictive streams, an environmental specialist, implemented as a graph neural network (GNN) that operationalizes the community as a directed contagion graph weighted by physics informed convection, radiation, and ember probabilities, and enriched with high dimensional Google AlphaEarth Foundation embeddings, and a Structural Specialist, implemented via XGBoost to isolate granular asset level resilience. Applied to the 2025 Eaton Fire, the framework reveals a critical dichotomy in risk drivers. The GNN demonstrates that neighborhood scale environmental pressure overwhelmingly dominates intrinsic structural features in defining propagation pathways, while the XGBoost model identifies eaves as the primary micro scale ingress vector. By synthesizing these divergent signals through logistic stacking, the ensemble achieves robust classification and generates a diagnostic risk topology. This capability empowers decision makers to move beyond binary loss prediction and precisely target mitigation prioritizing vegetation management for high connectivity clusters and structural hardening for architecturally vulnerable nodes thereby operationalizing a proactive, data driven approach to community resilience. --- ### Semantic Deception: When Reasoning Models Can't Compute an Addition Authors: Nathaniël de Leeuw, Marceau Nahon, Mathis Reymond, Raja Chatila, Mehdi Khamassi Published: 2025-12-23 Categories: cs.CL arXiv: https://arxiv.org/abs/2512.20812v1 Large language models (LLMs) are increasingly used in situations where human values are at stake, such as decision-making tasks that involve reasoning when performed by humans. We investigate the so-called reasoning capabilities of LLMs over novel symbolic representations by introducing an experimental framework that tests their ability to process and manipulate unfamiliar symbols. We introduce semantic deceptions: situations in which symbols carry misleading semantic associations due to their form, such as being embedded in specific contexts, designed to probe whether LLMs can maintain symbolic abstraction or whether they default to exploiting learned semantic associations. We redefine standard digits and mathematical operators using novel symbols, and task LLMs with solving simple calculations expressed in this altered notation. The objective is: (1) to assess LLMs' capacity for abstraction and manipulation of arbitrary symbol systems; (2) to evaluate their ability to resist misleading semantic cues that conflict with the task's symbolic logic. Through experiments with four LLMs we show that semantic cues can significantly deteriorate reasoning models' performance on very simple tasks. They reveal limitations in current LLMs' ability for symbolic manipulations and highlight a tendency to over-rely on surface-level semantics, suggesting that chain-of-thoughts may amplify reliance on statistical correlations. Even in situations where LLMs seem to correctly follow instructions, semantic cues still impact basic capabilities. These limitations raise ethical and societal concerns, undermining the widespread and pernicious tendency to attribute reasoning abilities to LLMs and suggesting how LLMs might fail, in particular in decision-making contexts where robust symbolic reasoning is essential and should not be compromised by residual semantic associations inherited from the model's training. --- ### Weighted MCC: A Robust Measure of Multiclass Classifier Performance for Observations with Individual Weights Authors: Rommel Cortez, Bala Krishnamoorthy Published: 2025-12-23 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2512.20811v1 Several performance measures are used to evaluate binary and multiclass classification tasks. But individual observations may often have distinct weights, and none of these measures are sensitive to such varying weights. We propose a new weighted Pearson-Matthews Correlation Coefficient (MCC) for binary classification as well as weighted versions of related multiclass measures. The weighted MCC varies between $-1$ and $1$. But crucially, the weighted MCC values are higher for classifiers that perform better on highly weighted observations, and hence is able to distinguish them from classifiers that have a similar overall performance and ones that perform better on the lowly weighted observations. Furthermore, we prove that the weighted measures are robust with respect to the choice of weights in a precise manner: if the weights are changed by at most $ε$, the value of the weighted measure changes at most by a factor of $ε$ in the binary case and by a factor of $ε^2$ in the multiclass case. Our computations demonstrate that the weighted measures clearly identify classifiers that perform better on higher weighted observations, while the unweighted measures remain completely indifferent to the choices of weights. --- ### Safety Alignment of LMs via Non-cooperative Games Authors: Anselm Paulus, Ilia Kulikov, Brandon Amos, Rémi Munos, Ivan Evtimov, Kamalika Chaudhuri, Arman Zharmagambetov Published: 2025-12-23 Categories: cs.AI arXiv: https://arxiv.org/abs/2512.20806v3 Ensuring the safety of language models (LMs) while maintaining their usefulness remains a critical challenge in AI alignment. Current approaches rely on sequential adversarial training: generating adversarial prompts and fine-tuning LMs to defend against them. We introduce a different paradigm: framing safety alignment as a non-zero-sum game between an Attacker LM and a Defender LM trained jointly via online reinforcement learning. Each LM continuously adapts to the other's evolving strategies, driving iterative improvement. Our method uses a preference-based reward signal derived from pairwise comparisons instead of point-wise scores, providing more robust supervision and potentially reducing reward hacking. Our RL recipe, AdvGame, shifts the Pareto frontier of safety and utility, yielding a Defender LM that is simultaneously more helpful and more resilient to adversarial attacks. In addition, the resulting Attacker LM converges into a strong, general-purpose red-teaming agent that can be directly deployed to probe arbitrary target models. Code at github.com/facebookresearch/advgame. --- ### A Benchmark for Evaluating Outcome-Driven Constraint Violations in Autonomous AI Agents Authors: Miles Q. Li, Benjamin C. M. Fung, Martin Weiss, Pulei Xiong, Khalil Al-Hussaeni, Claude Fachkha Published: 2025-12-23 Categories: cs.AI arXiv: https://arxiv.org/abs/2512.20798v5 As autonomous AI agents are increasingly deployed in high-stakes environments, ensuring their safety and alignment with human values is becoming a practical deployment concern. Current benchmarks for AI agents primarily evaluate refusal of explicitly harmful instructions or completion of complex multi-step tasks. However, there is a lack of benchmarks designed to capture emergent outcome-driven constraint violations, which arise when agents pursue goal optimization under strong performance incentives while deprioritizing ethical, legal, or safety constraints. To address this gap, we introduce a benchmark of 40 scenarios in production-inspired sandbox environments. Each scenario requires multi-step actions, and the agent's performance is tied to a specific Key Performance Indicator (KPI). Each scenario features Mandated (direct KPI-outcome mandate) and Incentivized (KPI-pressure-driven) variations to distinguish failures under direct outcome mandates from self-directed constraint violations. Across 12 state-of-the-art LLMs, we observe outcome-driven constraint violations ranging from 0.0% to 62.8%, with most evaluated models exhibiting misalignment rates at or above 25%. Furthermore, through a cross-generational analysis comparing current models with their predecessors within the same product families, we find that safety does not reliably improve across generations: misalignment rates rose in four families and fell in five. To improve evaluation robustness, we score trajectories with a four-model judge panel aggregated by median, finding high agreement on the primary misalignment threshold. We also observe substantial deliberative misalignment: cases where models later judge their own trajectories as unethical despite having executed them under KPI pressure. --- ### Measuring Mechanistic Independence: Can Bias Be Removed Without Erasing Demographics? Authors: Zhengyang Shan, Aaron Mueller Published: 2025-12-23 Categories: cs.CL arXiv: https://arxiv.org/abs/2512.20796v1 We investigate how independent demographic bias mechanisms are from general demographic recognition in language models. Using a multi-task evaluation setup where demographics are associated with names, professions, and education levels, we measure whether models can be debiased while preserving demographic detection capabilities. We compare attribution-based and correlation-based methods for locating bias features. We find that targeted sparse autoencoder feature ablations in Gemma-2-9B reduce bias without degrading recognition performance: attribution-based ablations mitigate race and gender profession stereotypes while preserving name recognition accuracy, whereas correlation-based ablations are more effective for education bias. Qualitative analysis further reveals that removing attribution features in education tasks induces ``prior collapse'', thus increasing overall bias. This highlights the need for dimension-specific interventions. Overall, our results show that demographic bias arises from task-specific mechanisms rather than absolute demographic markers, and that mechanistic inference-time interventions can enable surgical debiasing without compromising core model capabilities. --- ### Investigating Model Editing for Unlearning in Large Language Models Authors: Shariqah Hossain, Lalana Kagal Published: 2025-12-23 Categories: cs.CL arXiv: https://arxiv.org/abs/2512.20794v1 Machine unlearning aims to remove unwanted information from a model, but many methods are inefficient for LLMs with large numbers of parameters or fail to fully remove the intended information without degrading performance on knowledge that should be retained. Model editing algorithms solve a similar problem of changing information in models, but they focus on redirecting inputs to a new target rather than removing that information altogether. In this work, we explore the editing algorithms ROME, IKE, and WISE and design new editing targets for an unlearning setting. Through this investigation, we show that model editing approaches can exceed baseline unlearning methods in terms of quality of forgetting depending on the setting. Like traditional unlearning techniques, they struggle to encapsulate the scope of what is to be unlearned without damage to the overall model performance. --- ### X-GridAgent: An LLM-Powered Agentic AI System for Assisting Power Grid Analysis Authors: Yihan, Wen, Xin Chen Published: 2025-12-23 Categories: eess.SY, cs.AI arXiv: https://arxiv.org/abs/2512.20789v1 The growing complexity of power system operations has created an urgent need for intelligent, automated tools to support reliable and efficient grid management. Conventional analysis tools often require significant domain expertise and manual effort, which limits their accessibility and adaptability. To address these challenges, this paper presents X-GridAgent, a novel large language model (LLM)-powered agentic AI system designed to automate complex power system analysis through natural language queries. The system integrates domain-specific tools and specialized databases under a three-layer hierarchical architecture comprising planning, coordination, and action layers. This architecture offers high flexibility and adaptability to previously unseen tasks, while providing a modular and extensible framework that can be readily expanded to incorporate new tools, data sources, or analytical capabilities. To further enhance performance, we introduce two novel algorithms: (1) LLM-driven prompt refinement with human feedback, and (2) schema-adaptive hybrid retrieval-augmented generation (RAG) for accurate information retrieval from large-scale structured grid datasets. Experimental evaluations across a variety of user queries and power grid cases demonstrate the effectiveness and reliability of X-GridAgent in automating interpretable and rigorous power system analysis. --- ### Symbolic regression for defect interactions in 2D materials Authors: Mikhail Lazarev, Andrey Ustyuzhanin Published: 2025-12-23 Categories: cs.LG, cond-mat.mtrl-sci arXiv: https://arxiv.org/abs/2512.20785v2 Machine learning models have become firmly established across all scientific fields. Extracting features from data and making inferences based on them with neural network models often yields high accuracy; however, this approach has several drawbacks. Symbolic regression is a powerful technique for discovering analytical equations that describe data, providing interpretable and generalizable models capable of predicting unseen data. Symbolic regression methods have gained new momentum with the advancement of neural network technologies and offer several advantages, the main one being the interpretability of results. In this work, we examined the application of the deep symbolic regression algorithm SEGVAE to determine the properties of two-dimensional materials with defects. Comparing the results with state-of-the-art graph neural network-based methods shows comparable or, in some cases, even identical outcomes. We also discuss the applicability of this class of methods in natural sciences. --- ### NULLBUS: Multimodal Mixed-Supervision for Breast Ultrasound Segmentation via Nullable Global-Local Prompts Authors: Raja Mallina, Bryar Shareef Published: 2025-12-23 Categories: cs.CV, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2512.20783v1 Breast ultrasound (BUS) segmentation provides lesion boundaries essential for computer-aided diagnosis and treatment planning. While promptable methods can improve segmentation performance and tumor delineation when text or spatial prompts are available, many public BUS datasets lack reliable metadata or reports, constraining training to small multimodal subsets and reducing robustness. We propose NullBUS, a multimodal mixed-supervision framework that learns from images with and without prompts in a single model. To handle missing text, we introduce nullable prompts, implemented as learnable null embeddings with presence masks, enabling fallback to image-only evidence when metadata are absent and the use of text when present. Evaluated on a unified pool of three public BUS datasets, NullBUS achieves a mean IoU of 0.8568 and a mean Dice of 0.9103, demonstrating state-of-the-art performance under mixed prompt availability. --- ### Large Language Models Approach Expert Pedagogical Quality in Math Tutoring but Differ in Instructional and Linguistic Profiles Authors: Ramatu Oiza Abdulsalam, Segun Aroyehun Published: 2025-12-23 Categories: cs.CL, cs.CY arXiv: https://arxiv.org/abs/2512.20780v3 Recent work has explored the use of large language models (LLMs) to generate tutoring responses in mathematics, yet it remains unclear how closely their instructional behavior aligns with expert human practice. We analyze a dataset of math remediation dialogues in which expert tutors, novice tutors, and seven LLMs of varying sizes, comprising both open-weight and commercial models, respond to the same student errors. We examine instructional strategies and linguistic characteristics of tutoring responses, including uptake (restating and revoicing), pressing for accuracy and reasoning, lexical diversity, readability, politeness, and agency. We find that expert tutors produce higher-quality responses than novices, and that larger LLMs generally receive higher pedagogical quality ratings than smaller models, approaching expert performance on average. However, LLMs exhibit systematic differences in their instructional profiles: they underuse discursive strategies characteristic of expert tutors while generating longer, more lexically diverse, and more polite responses. Regression analyses show that pressing for accuracy and reasoning, restating and revoicing, and lexical diversity, are positively associated with perceived pedagogical quality, whereas higher levels of agentic and polite language are negatively associated. These findings highlight the importance of analyzing instructional strategies and linguistic characteristics when evaluating tutoring responses across human tutors and intelligent tutoring systems. --- ### Towards Optimal Performance and Action Consistency Guarantees in Dec-POMDPs with Inconsistent Beliefs and Limited Communication Authors: Moshe Rafaeli Shimron, Vadim Indelman Published: 2025-12-23 Categories: cs.MA, cs.AI, cs.RO arXiv: https://arxiv.org/abs/2512.20778v1 Multi-agent decision-making under uncertainty is fundamental for effective and safe autonomous operation. In many real-world scenarios, each agent maintains its own belief over the environment and must plan actions accordingly. However, most existing approaches assume that all agents have identical beliefs at planning time, implying these beliefs are conditioned on the same data. Such an assumption is often impractical due to limited communication. In reality, agents frequently operate with inconsistent beliefs, which can lead to poor coordination and suboptimal, potentially unsafe, performance. In this paper, we address this critical challenge by introducing a novel decentralized framework for optimal joint action selection that explicitly accounts for belief inconsistencies. Our approach provides probabilistic guarantees for both action consistency and performance with respect to open-loop multi-agent POMDP (which assumes all data is always communicated), and selectively triggers communication only when needed. Furthermore, we address another key aspect of whether, given a chosen joint action, the agents should share data to improve expected performance in inference. Simulation results show our approach outperforms state-of-the-art algorithms. --- ### Improving Matrix Exponential for Generative AI Flows: A Taylor-Based Approach Beyond Paterson--Stockmeyer Authors: Jorge Sastre, Daniel Faronbi, José Miguel Alonso, Peter Traver, Javier Ibáñez, Nuria Lloret Published: 2025-12-23 Categories: cs.LG, math.NA arXiv: https://arxiv.org/abs/2512.20777v2 The matrix exponential is a fundamental operator in scientific computing and system simulation, with applications ranging from control theory and quantum mechanics to modern generative machine learning. While Padé approximants combined with scaling and squaring have long served as the standard, recent Taylor-based methods, which utilize polynomial evaluation schemes that surpass the classical Paterson--Stockmeyer technique, offer superior accuracy and reduced computational complexity. This paper presents an optimized Taylor-based algorithm for the matrix exponential, specifically designed for the high-throughput requirements of generative AI flows. We provide a rigorous error analysis and develop a dynamic selection strategy for the Taylor order and scaling factor to minimize computational effort under a prescribed error tolerance. Extensive numerical experiments demonstrate that our approach provides significant acceleration and maintains high numerical stability compared to existing state-of-the-art implementations. These results establish the proposed method as a highly efficient tool for large-scale generative modeling. --- ### DIAL: Direct Iterative Adversarial Learning for Realistic Multi-Turn Dialogue Simulation Authors: Ziyi Zhu, Olivier Tieleman, Caitlin A. Stamatis, Luka Smyth, Thomas D. Hull, Daniel R. Cahn, Jinghong Chen, Matteo Malgaroli Published: 2025-12-23 Categories: cs.CL arXiv: https://arxiv.org/abs/2512.20773v4 Realistic user simulation is crucial for training and evaluating multi-turn dialogue systems, yet creating simulators that accurately replicate human behavior remains a significant challenge. An effective simulator must expose the failure modes of the systems under evaluation. This work introduces Direct Iterative Adversarial Learning (DIAL), an adversarial framework that iteratively enhances user simulator realism through a competitive dynamic between a generator (user simulator) and a discriminator. When applied to mental health support, a domain characterized by diverse failure types and a critical dependence on realistic user behavior for failure detection, DIAL restores lexical diversity diminished by supervised fine-tuning and drastically reduces discriminator accuracy. The resulting simulator exhibits a strong correlation between simulated and real failure occurrence rates while maintaining low distributional divergence of failure modes. These findings indicate that DIAL is a promising method for developing realistic user simulators in multi-turn dialogue, facilitating reliable and cost-effective system evaluation prior to deployment. --- ### OccuFly: A 3D Vision Benchmark for Semantic Scene Completion from the Aerial Perspective Authors: Markus Gross, Sai B. Matha, Aya Fahmy, Rui Song, Daniel Cremers, Henri Meess Published: 2025-12-23 Categories: cs.CV arXiv: https://arxiv.org/abs/2512.20770v2 Semantic Scene Completion (SSC) is essential for 3D perception in mobile robotics, as it enables holistic scene understanding by jointly estimating dense volumetric occupancy and per-voxel semantics. Although SSC has been widely studied in terrestrial domains such as autonomous driving, aerial settings like autonomous flying remain largely unexplored, thereby limiting progress on downstream applications. Furthermore, LiDAR sensors are the primary modality for SSC data generation, which poses challenges for most uncrewed aerial vehicles (UAVs) due to flight regulations, mass and energy constraints, and the sparsity of LiDAR point clouds from elevated viewpoints. To address these limitations, we propose a LiDAR-free, camera-based data generation framework. By leveraging classical 3D reconstruction, our framework automates semantic label transfer by lifting <10% of annotated images into the reconstructed point cloud, substantially minimizing manual 3D annotation effort. Based on this framework, we introduce OccuFly, the first real-world, camera-based aerial SSC benchmark, captured across multiple altitudes and all seasons. OccuFly provides over 20,000 samples of images, semantic voxel grids, and metric depth maps across 21 semantic classes in urban, industrial, and rural environments, and follows established data organization for seamless integration. We benchmark both SSC and metric monocular depth estimation on OccuFly, revealing fundamental limitations of current vision foundation models in aerial settings and establishing new challenges for robust 3D scene understanding in the aerial domain. Visit https://github.com/markus-42/occufly. --- ### Subgroup Discovery with the Cox Model Authors: Zachary Izzo, Iain Melvin Published: 2025-12-23 Categories: cs.LG, math.ST, stat.ML arXiv: https://arxiv.org/abs/2512.20762v2 We study the problem of subgroup discovery for survival analysis, where the goal is to find an interpretable subset of the data on which a Cox model is highly accurate. Our work is the first to study this particular subgroup problem, for which we make several contributions. Subgroup discovery methods generally require a "quality function" in order to sift through and select the most advantageous subgroups. We first examine why existing natural choices for quality functions are insufficient to solve the subgroup discovery problem for the Cox model. To address the shortcomings of existing metrics, we introduce two technical innovations: the *expected prediction entropy (EPE)*, a novel metric for evaluating survival models which predict a hazard function; and the *conditional rank statistics (CRS)*, a statistical object which quantifies the deviation of an individual point to the distribution of survival times in an existing subgroup. We study the EPE and CRS theoretically and show that they can solve many of the problems with existing metrics. We introduce a total of eight algorithms for the Cox subgroup discovery problem. The main algorithm is able to take advantage of both the EPE and the CRS, allowing us to give theoretical correctness results for this algorithm in a well-specified setting. We evaluate all of the proposed methods empirically on both synthetic and real data. The experiments confirm our theory, showing that our contributions allow for the recovery of a ground-truth subgroup in well-specified cases, as well as leading to better model fit compared to naively fitting the Cox model to the whole dataset in practical settings. Lastly, we conduct a case study on jet engine simulation data from NASA. The discovered subgroups uncover known nonlinearities/homogeneity in the data, and which suggest design choices which have been mirrored in practice. --- ### TS-Arena -- A Live Forecast Pre-Registration Platform Authors: Marcel Meyer, Sascha Kaltenpoth, Henrik Albers, Kevin Zalipski, Oliver Müller Published: 2025-12-23 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2512.20761v3 Time Series Foundation Models (TSFMs) are transforming the field of forecasting. However, evaluating them on historical data is increasingly difficult due to the risks of train-test sample overlaps and temporal overlaps between correlated train and test time series. To address this, we introduce TS-Arena, a live forecasting platform that shifts evaluation from the known past to the unknown future. Building on the concept of continuous benchmarking, TS-Arena evaluates models on future data. Crucially, we introduce a strict forecasting pre-registration protocol: models must submit predictions before the ground-truth data physically exists. This makes test-set contamination impossible by design. The platform relies on a modular microservice architecture that harmonizes and structures data from different sources and orchestrates containerized model submissions. By enforcing a strict pre-registration protocol on live data streams, TS-Arena prevents information leakage offers a faster alternative to traditional static, infrequently repeated competitions (e.g. the M-Competitions). First empirical results derived from operating TS-Arena over one year of energy time series demonstrate that established TSFMs accumulate robust longitudinal scores over time, while the continuous nature of the benchmark simultaneously allows newcomers to demonstrate immediate competitiveness. TS-Arena provides the necessary infrastructure to assess the true generalization capabilities of modern forecasting models. The platform and corresponding code are available at https://ts-arena.live/. --- ### Generalization of RLVR Using Causal Reasoning as a Testbed Authors: Brian Lu, Hongyu Zhao, Shuo Sun, Hao Peng, Rui Ding, Hongyuan Mei Published: 2025-12-23 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2512.20760v2 Reinforcement learning with verifiable rewards (RLVR) has emerged as a promising paradigm for post-training large language models (LLMs) on complex reasoning tasks. Yet, the conditions under which RLVR yields robust generalization remain underexplored. This paper provides an empirical study of RLVR generalization in the setting of probabilistic inference over causal graphical models. This setting offers two natural axes along which to examine generalization: (i) the level of the probabilistic query -- associational, interventional, or counterfactual -- and (ii) the structural complexity of the query, measured by the size of its relevant subgraph. We construct a dataset of causal graphs and queries spanning these difficulty axes and fine-tune Qwen-2.5-Instruct models using RLVR or supervised fine-tuning (SFT). We vary both the model scale (3B-32B) and the query level included in training. We find that RLVR yields stronger within-level and across-level generalization than SFT, but only for specific combinations of model size and training query level. Further analysis shows that RLVR's effectiveness depends on the model's initial reasoning competence. With sufficient initial competence, RLVR improves an LLM's marginalization strategy and reduces errors in intermediate probability calculations, producing substantial accuracy gains, particularly on more complex queries. These results show that RLVR can improve specific causal reasoning subskills, with its benefits emerging only when the model has sufficient initial competence. Our code and data is available at https://github.com/zhichul/rlcausal. --- ### TokSuite: Measuring the Impact of Tokenizer Choice on Language Model Behavior Authors: Gül Sena Altıntaş, Malikeh Ehghaghi, Brian Lester, Fengyuan Liu, Wanru Zhao, Marco Ciccone, Colin Raffel Published: 2025-12-23 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2512.20757v2 Tokenizers provide the fundamental basis through which text is represented and processed by language models (LMs). Despite the importance of tokenization, its role in LM performance and behavior is poorly understood due to the challenge of measuring the impact of tokenization in isolation. To address this need, we present TokSuite, a collection of models and a benchmark that supports research into tokenization's influence on LMs. Specifically, we release fourteen pre-trained models that use different off-the-shelf tokenizers but are otherwise identical, using the same architecture, dataset, training budget, and initialization. We also release a multilingual robustness benchmark that measures model performance under real-world perturbations in English, Chinese, Farsi, Italian, and Turkish, curated by native annotators. Together, TokSuite allows robust decoupling of the influence of a model's tokenizer, supporting a series of novel findings that elucidate the respective benefits and shortcomings of a wide range of popular tokenizers. --- ### Bridging Efficiency and Safety: Formal Verification of Neural Networks with Early Exits Authors: Yizhak Yisrael Elboher, Avraham Raviv, Amihay Elboher, Zhouxing Shi, Omri Azencot, Hillel Kugler, Guy Katz Published: 2025-12-23 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2512.20755v1 Ensuring the safety and efficiency of AI systems is a central goal of modern research. Formal verification provides guarantees of neural network robustness, while early exits improve inference efficiency by enabling intermediate predictions. Yet verifying networks with early exits introduces new challenges due to their conditional execution paths. In this work, we define a robustness property tailored to early exit architectures and show how off-the-shelf solvers can be used to assess it. We present a baseline algorithm, enhanced with an early stopping strategy and heuristic optimizations that maintain soundness and completeness. Experiments on multiple benchmarks validate our framework's effectiveness and demonstrate the performance gains of the improved algorithm. Alongside the natural inference acceleration provided by early exits, we show that they also enhance verifiability, enabling more queries to be solved in less time compared to standard networks. Together with a robustness analysis, we show how these metrics can help users navigate the inherent trade-off between accuracy and efficiency. ---