# Research Papers — 2026-01 · bundle page 82 of 189 > Full-text markdown of papers #4051–4100 of 9403 published in 2026-01, newest first. AI-readable mirrors of the full 9403-paper 2026-01 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-01/page/82 [← Previous bundle](https://aipod.dev/papers/archive/2026-01/page/81.md) [Next bundle →](https://aipod.dev/papers/archive/2026-01/page/83.md) ### Unsupervised Video Class-Incremental Learning via Deep Embedded Clustering Management Authors: Nattapong Kurpukdee, Adrian G. Bors Published: 2026-01-20 Categories: cs.CV, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2601.14069v1 Unsupervised video class incremental learning (uVCIL) represents an important learning paradigm for learning video information without forgetting, and without considering any data labels. Prior approaches have focused on supervised class-incremental learning, relying on using the knowledge of labels and task boundaries, which is costly, requires human annotation, or is simply not a realistic option. In this paper, we propose a simple yet effective approach to address the uVCIL. We first consider a deep feature extractor network, providing a set of representative video features during each task without assuming any class or task information. We then progressively build a series of deep clusters from the extracted features. During the successive task learning, the model updated from the previous task is used as an initial state in order to transfer knowledge to the current learning task. We perform in-depth evaluations on three standard video action recognition datasets, including UCF101, HMDB51, and Something-to-Something V2, by ignoring the labels from the supervised setting. Our approach significantly outperforms other baselines on all datasets. --- ### VERIDAH: Solving Enumeration Anomaly Aware Vertebra Labeling across Imaging Sequences Authors: Hendrik Möller, Hanna Schoen, Robert Graf, Matan Atad, Nathan Molinier, Anjany Sekuboyina, Bettina K. Budai, Fabian Bamberg, Steffen Ringhof, Christopher Schlett, Tobias Pischon, Thoralf Niendorf et al. Published: 2026-01-20 Categories: cs.CV arXiv: https://arxiv.org/abs/2601.14066v1 The human spine commonly consists of seven cervical, twelve thoracic, and five lumbar vertebrae. However, enumeration anomalies may result in individuals having eleven or thirteen thoracic vertebrae and four or six lumbar vertebrae. Although the identification of enumeration anomalies has potential clinical implications for chronic back pain and operation planning, the thoracolumbar junction is often poorly assessed and rarely described in clinical reports. Additionally, even though multiple deep-learning-based vertebra labeling algorithms exist, there is a lack of methods to automatically label enumeration anomalies. Our work closes that gap by introducing "Vertebra Identification with Anomaly Handling" (VERIDAH), a novel vertebra labeling algorithm based on multiple classification heads combined with a weighted vertebra sequence prediction algorithm. We show that our approach surpasses existing models on T2w TSE sagittal (98.30% vs. 94.24% of subjects with all vertebrae correctly labeled, p < 0.001) and CT imaging (99.18% vs. 77.26% of subjects with all vertebrae correctly labeled, p < 0.001) and works in arbitrary field-of-view images. VERIDAH correctly labeled the presence 2 Möller et al. of thoracic enumeration anomalies in 87.80% and 96.30% of T2w and CT images, respectively, and lumbar enumeration anomalies in 94.48% and 97.22% for T2w and CT, respectively. Our code and models are available at: https://github.com/Hendrik-code/spineps. --- ### XCR-Bench: Benchmarking Cross-Cultural Reasoning in LLMs via Culture-Specific Items and Hall's Triad Authors: Mohsinul Kabir, Tasnim Ahmed, Md Mezbaur Rahman, Shaoxiong Ji, Hassan Alhuzali, Yuechen Jiang, Jimin Huang, Sophia Ananiadou Published: 2026-01-20 Categories: cs.CL, cs.AI, cs.CY arXiv: https://arxiv.org/abs/2601.14063v2 Cross-cultural competence in large language models (LLMs) requires understanding and adapting Culture-Specific Items (CSIs) across varying cultural contexts. However, progress in evaluating this capability remains limited by the lack of high-quality CSI-annotated corpora with parallel cross-cultural sentence pairs. We introduce XCR-Bench, a Cross(X)-Cultural Reasoning Benchmark containing 4.1k parallel sentences and 1,098 CSIs across three reasoning tasks. XCR-Bench integrates Newmark's CSI framework with Hall's Triad of Culture, enabling evaluation across levels of cultural visibility -- from observable practices to implicit social norms and values. Experiments on eight multilingual LLMs show that state-of-the-art models exhibit consistent weaknesses in identifying and adapting specific categories of CSIs, revealing a gap between surface-level recall and explicit cultural reasoning. Performance declines significantly on culturally sensitive categories and deeper cultural levels (p<0.005, 8/8 models), and adaptation quality varies systematically across target cultures and Bengali regional variants, indicating encoded regional and ethno-religious biases even within a single linguistic setting. We publicly release the corpus and code to support future research on cross-cultural NLP. --- ### Demystifying the trend of the healthcare index: Is historical price a key driver? Authors: Payel Sadhukhan, Samrat Gupta, Subhasis Ghosh, Tanujit Chakraborty Published: 2026-01-20 Categories: q-fin.ST, stat.AP, stat.ML arXiv: https://arxiv.org/abs/2601.14062v1 Healthcare sector indices consolidate the economic health of pharmaceutical, biotechnology, and healthcare service firms. The short-term movements in these indices are closely intertwined with capital allocation decisions affecting research and development investment, drug availability, and long-term health outcomes. This research investigates whether historical open-high-low-close (OHLC) index data contain sufficient information for predicting the directional movement of the opening index on the subsequent trading day. The problem is formulated as a supervised classification task involving a one-step-ahead rolling window. A diverse feature set is constructed, comprising original prices, volatility-based technical indicators, and a novel class of nowcasting features derived from mutual OHLC ratios. The framework is evaluated on data from healthcare indices in the U.S. and Indian markets over a five-year period spanning multiple economic phases, including the COVID-19 pandemic. The results demonstrate robust predictive performance, with accuracy exceeding 0.8 and Matthews correlation coefficients above 0.6. Notably, the proposed nowcasting features have emerged as a key determinant of the market movement. We have employed the Shapley-based explainability paradigm to further elucidate the contribution of the features: outcomes reveal the dominant role of the nowcasting features, followed by a more moderate contribution of original prices. This research offers a societal utility: the proposed features and model for short-term forecasting of healthcare indices can reduce information asymmetry and support a more stable and equitable health economy. --- ### Fine-Grained Zero-Shot Composed Image Retrieval with Complementary Visual-Semantic Integration Authors: Yongcong Ye, Kai Zhang, Yanghai Zhang, Enhong Chen, Longfei Li, Jun Zhou Published: 2026-01-20 Categories: cs.CV arXiv: https://arxiv.org/abs/2601.14060v1 Zero-shot composed image retrieval (ZS-CIR) is a rapidly growing area with significant practical applications, allowing users to retrieve a target image by providing a reference image and a relative caption describing the desired modifications. Existing ZS-CIR methods often struggle to capture fine-grained changes and integrate visual and semantic information effectively. They primarily rely on either transforming the multimodal query into a single text using image-to-text models or employing large language models for target image description generation, approaches that often fail to capture complementary visual information and complete semantic context. To address these limitations, we propose a novel Fine-Grained Zero-Shot Composed Image Retrieval method with Complementary Visual-Semantic Integration (CVSI). Specifically, CVSI leverages three key components: (1) Visual Information Extraction, which not only extracts global image features but also uses a pre-trained mapping network to convert the image into a pseudo token, combining it with the modification text and the objects most likely to be added. (2) Semantic Information Extraction, which involves using a pre-trained captioning model to generate multiple captions for the reference image, followed by leveraging an LLM to generate the modified captions and the objects most likely to be added. (3) Complementary Information Retrieval, which integrates information extracted from both the query and database images to retrieve the target image, enabling the system to efficiently handle retrieval queries in a variety of situations. Extensive experiments on three public datasets (e.g., CIRR, CIRCO, and FashionIQ) demonstrate that CVSI significantly outperforms existing state-of-the-art methods. Our code is available at https://github.com/yyc6631/CVSI. --- ### POCI-Diff: 3D-Layout Guided Diffusion for Controllable Synthetic Surveillance Data Generation Authors: Andrea Rigo, Luca Stornaiuolo, Weijie Wang, Mauro Martino, Bruno Lepri, Nicu Sebe Published: 2026-01-20 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2601.14056v2 Training robust visual surveillance models requires large-scale datasets with precise spatial annotations, yet collecting real surveillance data is costly, privacy-sensitive, and often legally constrained. Synthetic data generation offers a compelling alternative, but existing methods lack fine-grained 3D control over object placement and appearance, limiting geometric consistency across camera viewpoints. We introduce POCI-Diff (Positioning Objects Consistently and Interactively), a framework that generates annotated synthetic scenes from explicit 3D bounding-box layouts with per-object semantic control. By integrating Blended Latent Diffusion with depth-conditioned ControlNet, POCI-Diff synthesises complex multi-object scenes in a single forward pass, binding individual text descriptions to specific 3D locations. We further propose a warping-free editing pipeline supporting object insertion, removal, and transformation via regeneration, enabling efficient scene variation for data augmentation. Object identity across edits is preserved by conditioning on reference images via IP-Adapter, ensuring appearance consistency throughout interactive scene manipulation. Experiments show that POCI-Diff outperforms state-of-the-art 3D layout-guided generation methods in visual fidelity and layout adherence, while eliminating warping-induced geometric artifacts. --- ### Decoder-Free Supervoxel GNN for Accurate Brain-Tumor Localization in Multi-Modal MRI Authors: Andrea Protani, Marc Molina Van Den Bosch, Lorenzo Giusti, Heloisa Barbosa Da Silva, Paolo Cacace, Albert Sund Aillet, Miguel Angel Gonzalez Ballester, Friedhelm Hummel, Luigi Serio Published: 2026-01-20 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2601.14055v1 Modern vision backbones for 3D medical imaging typically process dense voxel grids through parameter-heavy encoder-decoder structures, a design that allocates a significant portion of its parameters to spatial reconstruction rather than feature learning. Our approach introduces SVGFormer, a decoder-free pipeline built upon a content-aware grouping stage that partitions the volume into a semantic graph of supervoxels. Its hierarchical encoder learns rich node representations by combining a patch-level Transformer with a supervoxel-level Graph Attention Network, jointly modeling fine-grained intra-region features and broader inter-regional dependencies. This design concentrates all learnable capacity on feature encoding and provides inherent, dual-scale explainability from the patch to the region level. To validate the framework's flexibility, we trained two specialized models on the BraTS dataset: one for node-level classification and one for tumor proportion regression. Both models achieved strong performance, with the classification model achieving a F1-score of 0.875 and the regression model a MAE of 0.028, confirming the encoder's ability to learn discriminative and localized features. Our results establish that a graph-based, encoder-only paradigm offers an accurate and inherently interpretable alternative for 3D medical image representation. --- ### SecureSplit: Mitigating Backdoor Attacks in Split Learning Authors: Zhihao Dou, Dongfei Cui, Weida Wang, Anjun Gao, Yueyang Quan, Mengyao Ma, Viet Vo, Guangdong Bai, Zhuqing Liu, Minghong Fang Published: 2026-01-20 Categories: cs.CR, cs.DC, cs.LG arXiv: https://arxiv.org/abs/2601.14054v2 Split Learning (SL) offers a framework for collaborative model training that respects data privacy by allowing participants to share the same dataset while maintaining distinct feature sets. However, SL is susceptible to backdoor attacks, in which malicious clients subtly alter their embeddings to insert hidden triggers that compromise the final trained model. To address this vulnerability, we introduce SecureSplit, a defense mechanism tailored to SL. SecureSplit applies a dimensionality transformation strategy to accentuate subtle differences between benign and poisoned embeddings, facilitating their separation. With this enhanced distinction, we develop an adaptive filtering approach that uses a majority-based voting scheme to remove contaminated embeddings while preserving clean ones. Rigorous experiments across four datasets (CIFAR-10, MNIST, CINIC-10, and ImageNette), five backdoor attack scenarios, and seven alternative defenses confirm the effectiveness of SecureSplit under various challenging conditions. --- ### LLMOrbit: A Circular Taxonomy of Large Language Models -From Scaling Walls to Agentic AI Systems Authors: Badri N. Patro, Vijay S. Agneeswaran Published: 2026-01-20 Categories: cs.LG, cs.AI, cs.CV, cs.MA, eess.IV arXiv: https://arxiv.org/abs/2601.14053v2 The field of artificial intelligence has undergone a revolution from foundational Transformer architectures to reasoning-capable systems approaching human-level performance. We present LLMOrbit, a comprehensive circular taxonomy navigating the landscape of large language models spanning 2019-2025. This survey examines over 50 models across 15 organizations through eight interconnected orbital dimensions, documenting architectural innovations, training methodologies, and efficiency patterns defining modern LLMs, generative AI, and agentic systems. We identify three critical crises: (1) data scarcity (9-27T tokens depleted by 2026-2028), (2) exponential cost growth ($3M to $300M+ in 5 years), and (3) unsustainable energy consumption (22x increase), establishing the scaling wall limiting brute-force approaches. Our analysis reveals six paradigms breaking this wall: (1) test-time compute (o1, DeepSeek-R1 achieve GPT-4 performance with 10x inference compute), (2) quantization (4-8x compression), (3) distributed edge computing (10x cost reduction), (4) model merging, (5) efficient training (ORPO reduces memory 50%), and (6) small specialized models (Phi-4 14B matches larger models). Three paradigm shifts emerge: (1) post-training gains (RLHF, GRPO, pure RL contribute substantially, DeepSeek-R1 achieving 79.8% MATH), (2) efficiency revolution (MoE routing 18x efficiency, Multi-head Latent Attention 8x KV cache compression enables GPT-4-level performance at $<$$0.30/M tokens), and (3) democratization (open-source Llama 3 88.6% MMLU surpasses GPT-4 86.4%). We provide insights into techniques (RLHF, PPO, DPO, GRPO, ORPO), trace evolution from passive generation to tool-using agents (ReAct, RAG, multi-agent systems), and analyze post-training innovations. --- ### Vision Also You Need: Navigating Out-of-Distribution Detection with Multimodal Large Language Model Authors: Haoran Xu, Yanlin Liu, Zizhao Tong, Jiaze Li, Kexue Fu, Yuyang Zhang, Longxiang Gao, Shuaiguang Li, Xingyu Li, Yanran Xu, Changwei Wang Published: 2026-01-20 Categories: cs.CV arXiv: https://arxiv.org/abs/2601.14052v1 Out-of-Distribution (OOD) detection is a critical task that has garnered significant attention. The emergence of CLIP has spurred extensive research into zero-shot OOD detection, often employing a training-free approach. Current methods leverage expert knowledge from large language models (LLMs) to identify potential outliers. However, these approaches tend to over-rely on knowledge in the text space, neglecting the inherent challenges involved in detecting out-of-distribution samples in the image space. In this paper, we propose a novel pipeline, MM-OOD, which leverages the multimodal reasoning capabilities of MLLMs and their ability to conduct multi-round conversations for enhanced outlier detection. Our method is designed to improve performance in both near OOD and far OOD tasks. Specifically, (1) for near OOD tasks, we directly feed ID images and corresponding text prompts into MLLMs to identify potential outliers; and (2) for far OOD tasks, we introduce the sketch-generate-elaborate framework: first, we sketch outlier exposure using text prompts, then generate corresponding visual OOD samples, and finally elaborate by using multimodal prompts. Experiments demonstrate that our method achieves significant improvements on widely used multimodal datasets such as Food-101, while also validating its scalability on ImageNet-1K. --- ### Kakugo: Distillation of Low-Resource Languages into Small Language Models Authors: Peter Devine, Mardhiyah Sanni, Farid Adilazuarda, Julieta Gil Loizaga, Barry Haddow Published: 2026-01-20 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2601.14051v1 We present Kakugo, a novel and cost-effective pipeline designed to train general-purpose Small Language Models (SLMs) for low-resource languages using only the language name as input. By using a large teacher model to generate synthetic prompts and translate instruction datasets, we produced training data and SLMs for 54 low-resource languages. Evaluations across a diverse set of general natural language processing tasks, including translation, classification, and question answering, demonstrate that our pipeline consistently improves performance over base models. With a total generation and training cost of under $50 per language, Kakugo offers an accessible method for communities to develop language-specific AI. --- ### Understanding Multilingualism in Mixture-of-Experts LLMs: Routing Mechanism, Expert Specialization, and Layerwise Steering Authors: Yuxin Chen, Zhengzhou Cai, Xiangtian Ji, Weixiang Zhao, An Zhang, Xiang Wang, Tat-Seng Chua Published: 2026-01-20 Categories: cs.CL arXiv: https://arxiv.org/abs/2601.14050v1 Mixture-of-Experts (MoE) architectures have shown strong multilingual capabilities, yet the internal mechanisms underlying performance gains and cross-language differences remain insufficiently understood. In this work, we conduct a systematic analysis of MoE models, examining routing behavior and expert specialization across languages and network depth. Our analysis reveals that multilingual processing in MoE models is highly structured: routing aligns with linguistic families, expert utilization follows a clear layerwise pattern, and high-resource languages rely on shared experts while low-resource languages depend more on language-exclusive experts despite weaker performance. Layerwise interventions further show that early and late MoE layers support language-specific processing, whereas middle layers serve as language-agnostic capacity hubs. Building on these insights, we propose a routing-guided steering method that adaptively guides routing behavior in middle layers toward shared experts associated with dominant languages at inference time, leading to consistent multilingual performance improvements, particularly for linguistically related language pairs. Our code is available at https://github.com/conctsai/Multilingualism-in-Mixture-of-Experts-LLMs. --- ### Collective intelligence in science: direct elicitation of diverse information from experts with unknown information structure Authors: Alexey V. Osipov, Nikolay N. Osipov Published: 2026-01-20 Categories: cs.GT, cs.AI, cs.MA, cs.SI, econ.TH arXiv: https://arxiv.org/abs/2601.14047v2 Suppose we need a deep collective analysis of an open scientific problem: there is a complex scientific hypothesis and a large online group of mutually unrelated experts with relevant private information of a diverse and unpredictable nature. This information may be results of experts' individual experiments, original reasoning of some of them, results of AI systems they use, etc. We propose a simple mechanism based on a self-resolving play-money prediction market entangled with a chat. We show that such a system can easily be brought to an equilibrium where participants directly share their private information on the hypothesis through the chat and trade as if the market were resolved in accordance with the truth of the hypothesis. This approach will lead to efficient aggregation of relevant information in a completely interpretable form even if the ground truth cannot be established and experts initially know nothing about each other and cannot perform complex Bayesian calculations. Finally, by rewarding the experts with some real assets proportionally to the play money they end up with, we can get an innovative way to fund large-scale collaborative studies of any type. --- ### PRiSM: Benchmarking Phone Realization in Speech Models Authors: Shikhar Bharadwaj, Chin-Jou Li, Yoonjae Kim, Kwanghee Choi, Eunjung Yeo, Ryan Soh-Eun Shim, Hanyu Zhou, Brendon Boldt, Karen Rosero Jacome, Kalvin Chang, Darsh Agrawal, Keer Xu et al. Published: 2026-01-20 Categories: cs.CL, cs.SD arXiv: https://arxiv.org/abs/2601.14046v2 Phone recognition (PR) serves as the atomic interface for language-agnostic modeling for cross-lingual speech processing and phonetic analysis. Despite prolonged efforts in developing PR systems, current evaluations only measure surface-level transcription accuracy. We introduce PRiSM, the first open-source benchmark designed to expose blind spots in phonetic perception through intrinsic and extrinsic evaluation of PR systems. PRiSM standardizes transcription-based evaluation and assesses downstream utility in clinical, educational, and multilingual settings with transcription and representation probes. We find that diverse language exposure during training is key to PR performance, encoder-CTC models are the most stable, and specialized PR models still outperform Large Audio Language Models. PRiSM releases code, recipes, and datasets to move the field toward multilingual speech models with robust phonetic ability: https://github.com/changelinglab/prism. --- ### Weather-R1: Logically Consistent Reinforcement Fine-Tuning for Multimodal Reasoning in Meteorology Authors: Kaiyu Wu, Pucheng Han, Hualong Zhang, Naigeng Wu, Keze Wang Published: 2026-01-20 Categories: cs.CV arXiv: https://arxiv.org/abs/2601.14044v1 While Vision Language Models (VLMs) show advancing reasoning capabilities, their application in meteorology is constrained by a domain gap and a reasoning faithfulness gap. Specifically, mainstream Reinforcement Fine-Tuning (RFT) can induce Self-Contradictory Reasoning (Self-Contra), where the model's reasoning contradicts its final answer, which is unacceptable in such a high-stakes domain. To address these challenges, we construct WeatherQA, a novel multimodal reasoning benchmark in meteorology. We also propose Logically Consistent Reinforcement Fine-Tuning (LoCo-RFT), which resolves Self-Contra by introducing a logical consistency reward. Furthermore, we introduce Weather-R1, the first reasoning VLM with logical faithfulness in meteorology, to the best of our knowledge. Experiments demonstrate that Weather-R1 improves performance on WeatherQA by 9.8 percentage points over the baseline, outperforming Supervised Fine-Tuning and RFT, and even surpassing the original Qwen2.5-VL-32B. These results highlight the effectiveness of our LoCo-RFT and the superiority of Weather-R1. Our benchmark and code are available at https://github.com/Marcowky/Weather-R1. --- ### Federated Balanced Learning Authors: Jiaze Li, Haoran Xu, Wanyi Wu, Changwei Wang, Shuaiguang Li, Jianzhong Ju, Zhenbo Luo, Jian Luan, Youyang Qu, Longxiang Gao, Xudong Yang, Lumin Xing Published: 2026-01-20 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2601.14042v2 Federated learning is a paradigm of joint learning in which clients collaborate by sharing model parameters instead of data. However, in the non-iid setting, the global model experiences client drift, which can seriously affect the final performance of the model. Previous methods tend to correct the global model that has already deviated based on the loss function or gradient, overlooking the impact of the client samples. In this paper, we rethink the role of the client side and propose Federated Balanced Learning, i.e., FBL, to prevent this issue from the beginning through sample balance on the client side. Technically, FBL allows unbalanced data on the client side to achieve sample balance through knowledge filling and knowledge sampling using edge-side generation models, under the limitation of a fixed number of data samples on clients. Furthermore, we design a Knowledge Alignment Strategy to bridge the gap between synthetic and real data, and a Knowledge Drop Strategy to regularize our method. Meanwhile, we scale our method to real and complex scenarios, allowing different clients to adopt various methods, and extend our framework to further improve performance. Numerous experiments show that our method outperforms state-of-the-art baselines. The code is released upon acceptance. --- ### Top 10 Open Challenges Steering the Future of Diffusion Language Model and Its Variants Authors: Yunhe Wang, Kai Han, Huiling Zhen, Yuchuan Tian, Hanting Chen, Yongbing Huang, Yufei Cui, Yingte Shu, Shan Gao, Ismail Elezi, Roy Vaughan Miles, Songcen Xu et al. Published: 2026-01-20 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2601.14041v1 The paradigm of Large Language Models (LLMs) is currently defined by auto-regressive (AR) architectures, which generate text through a sequential ``brick-by-brick'' process. Despite their success, AR models are inherently constrained by a causal bottleneck that limits global structural foresight and iterative refinement. Diffusion Language Models (DLMs) offer a transformative alternative, conceptualizing text generation as a holistic, bidirectional denoising process akin to a sculptor refining a masterpiece. However, the potential of DLMs remains largely untapped as they are frequently confined within AR-legacy infrastructures and optimization frameworks. In this Perspective, we identify ten fundamental challenges ranging from architectural inertia and gradient sparsity to the limitations of linear reasoning that prevent DLMs from reaching their ``GPT-4 moment''. We propose a strategic roadmap organized into four pillars: foundational infrastructure, algorithmic optimization, cognitive reasoning, and unified multimodal intelligence. By shifting toward a diffusion-native ecosystem characterized by multi-scale tokenization, active remasking, and latent thinking, we can move beyond the constraints of the causal horizon. We argue that this transition is essential for developing next-generation AI capable of complex structural reasoning, dynamic self-correction, and seamless multimodal integration. --- ### Generalizing Abstention for Noise-Robust Learning in Medical Image Segmentation Authors: Wesam Moustafa, Hossam Elsafty, Helen Schneider, Lorenz Sparrenberg, Rafet Sifa Published: 2026-01-20 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2601.14039v1 Label noise is a critical problem in medical image segmentation, often arising from the inherent difficulty of manual annotation. Models trained on noisy data are prone to overfitting, which degrades their generalization performance. While a number of methods and strategies have been proposed to mitigate noisy labels in the segmentation domain, this area remains largely under-explored. The abstention mechanism has proven effective in classification tasks by enhancing the capabilities of Cross Entropy, yet its potential in segmentation remains unverified. In this paper, we address this gap by introducing a universal and modular abstention framework capable of enhancing the noise-robustness of a diverse range of loss functions. Our framework improves upon prior work with two key components: an informed regularization term to guide abstention behaviour, and a more flexible power-law-based auto-tuning algorithm for the abstention penalty. We demonstrate the framework's versatility by systematically integrating it with three distinct loss functions to create three novel, noise-robust variants: GAC, SAC, and ADS. Experiments on the CaDIS and DSAD medical datasets show our methods consistently and significantly outperform their non-abstaining baselines, especially under high noise levels. This work establishes that enabling models to selectively ignore corrupted samples is a powerful and generalizable strategy for building more reliable segmentation models. Our code is publicly available at https://github.com/wemous/abstention-for-segmentation. --- ### Correcting and Quantifying Systematic Errors in 3D Box Annotations for Autonomous Driving Authors: Alexandre Justo Miro, Ludvig af Klinteberg, Bogdan Timus, Aron Asefaw, Ajinkya Khoche, Thomas Gustafsson, Sina Sharif Mansouri, Masoud Daneshtalab Published: 2026-01-20 Categories: cs.CV arXiv: https://arxiv.org/abs/2601.14038v1 Accurate ground truth annotations are critical to supervised learning and evaluating the performance of autonomous vehicle systems. These vehicles are typically equipped with active sensors, such as LiDAR, which scan the environment in predefined patterns. 3D box annotation based on data from such sensors is challenging in dynamic scenarios, where objects are observed at different timestamps, hence different positions. Without proper handling of this phenomenon, systematic errors are prone to being introduced in the box annotations. Our work is the first to discover such annotation errors in widely used, publicly available datasets. Through our novel offline estimation method, we correct the annotations so that they follow physically feasible trajectories and achieve spatial and temporal consistency with the sensor data. For the first time, we define metrics for this problem; and we evaluate our method on the Argoverse 2, MAN TruckScenes, and our proprietary datasets. Our approach increases the quality of box annotations by more than 17% in these datasets. Furthermore, we quantify the annotation errors in them and find that the original annotations are misplaced by up to 2.5 m, with highly dynamic objects being the most affected. Finally, we test the impact of the errors in benchmarking and find that the impact is larger than the improvements that state-of-the-art methods typically achieve with respect to the previous state-of-the-art methods; showing that accurate annotations are essential for correct interpretation of performance. Our code is available at https://github.com/alexandre-justo-miro/annotation-correction-3D-boxes. --- ### Private Prediction via PAC Privacy Authors: Xiaochen Zhu, Mayuri Sridhar, Srinivas Devadas Published: 2026-01-20 Categories: cs.LG, cs.CR arXiv: https://arxiv.org/abs/2601.14033v2 Machine learning models are increasingly served behind APIs. This renders private prediction, i.e., privatizing a model's outputs rather than its parameters, a natural privacy target: model outputs are lower-dimensional and far more stable to training-data changes than weights. While differential privacy (DP) cannot effectively exploit this as it calibrates noise to worst-case sensitivity that is intractable to bound for non-convex models, we argue that PAC privacy is a natural fit for private prediction. It is instance-based, and calibrates noise to a black-box function's empirical stability to control mutual-information (MI) leakage. The missing ingredient is efficient, adaptive composition. Serving predictions means answering a long stream of adaptively chosen queries from untrusted users; existing composition either fails under adaptivity, grows quadratically, or reverts to input-independent, DP-like noise. We close this gap with a new adversarial composition result via adaptive noise calibration and prove that MI accumulates only linearly under adaptive and adversarial querying. Experiments across modalities show that prediction stability enables high utility even at a tiny per-query budget: on CIFAR-10, we achieve 87.79% accuracy with a per-query MI budget of $2^{-32}$. This enables serving one million queries while provably bounding membership-inference success to 51.08% -- the same guarantee as $(0.04, 10^{-5})$-DP. Further, in the presence of auxiliary public data, the large volume of PAC-private predictions enables us to distill a publishable model that can be queried without limit. Concretely, 210,000 private labels on an ImageNet subset distill into a student reaching 91.86% accuracy on CIFAR-10 with membership inference success bounded by 50.49%, comparable to $(0.02, 10^{-5})$-DP. --- ### RM-Distiller: Exploiting Generative LLM for Reward Model Distillation Authors: Hongli Zhou, Hui Huang, Wei Liu, Chenglong Wang, Xingyuan Bu, Lvyuan Han, Fuhai Song, Muyun Yang, Wenhao Jiang, Hailong Cao, Tiejun Zhao Published: 2026-01-20 Categories: cs.CL arXiv: https://arxiv.org/abs/2601.14032v2 Reward models (RMs) play a pivotal role in aligning large language models (LLMs) with human preferences. Due to the difficulty of obtaining high-quality human preference annotations, distilling preferences from generative LLMs has emerged as a standard practice. However, existing approaches predominantly treat teacher models as simple binary annotators, failing to fully exploit the rich knowledge and capabilities for RM distillation. To address this, we propose RM-Distiller, a framework designed to systematically exploit the multifaceted capabilities of teacher LLMs: (1) Refinement capability, which synthesizes highly correlated response pairs to create fine-grained and contrastive signals. (2) Scoring capability, which guides the RM in capturing precise preference strength via a margin-aware optimization objective. (3) Generation capability, which incorporates the teacher's generative distribution to regularize the RM to preserve its fundamental linguistic knowledge. Extensive experiments demonstrate that RM-Distiller significantly outperforms traditional distillation methods both on RM benchmarks and reinforcement learning-based alignment, proving that exploiting multifaceted teacher capabilities is critical for effective reward modeling. To the best of our knowledge, this is the first systematic research on RM distillation from generative LLMs. --- ### Intermittent time series forecasting: local vs global models Authors: Stefano Damato, Nicolò Rubattu, Dario Azzimonti, Giorgio Corani Published: 2026-01-20 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2601.14031v2 Forecasting intermittent time series, which contain zeros, is a crucial challenge in supply chains as inventory policies require probabilistic forecasts to establish safety levels. Intermittent time series are commonly forecast using local models, trained individually on each time series. In the last years global models, trained on a large collection of time series, have become popular for time series forecasting. Global models are often based on neural networks or gradient boosted trees. We carry out the first study comparing state-of-the-art probabilistic local and global models on intermittent time series. For global models we consider three different distribution heads suitable for intermittent time series: negative binomial, hurdle-shifted negative binomial and Tweedie. To the best of our knowledge, this is the first use of the latter two with neural networks. We perform experiments on five datasets comprising overall more than 40'000 real-world time series. Among global models, TiDE, a simple neural network architecture, achieves the best accuracy; it also consistently outperforms local models and has lower computational requirements. Large global models are instead much more computationally demanding and less accurate. Among the distribution heads, the Tweedie provides the best estimates of the highest quantiles. --- ### Likelihood-Separable Diffusion Inference for Multi-Image MRI Super-Resolution Authors: Samuel W. Remedios, Zhangxing Bian, Shuwen Wei, Aaron Carass, Jerry L. Prince, Blake E. Dewey Published: 2026-01-20 Categories: cs.CV arXiv: https://arxiv.org/abs/2601.14030v1 Diffusion models are the current state-of-the-art for solving inverse problems in imaging. Their impressive generative capability allows them to approximate sampling from a prior distribution, which alongside a known likelihood function permits posterior sampling without retraining the model. While recent methods have made strides in advancing the accuracy of posterior sampling, the majority focuses on single-image inverse problems. However, for modalities such as magnetic resonance imaging (MRI), it is common to acquire multiple complementary measurements, each low-resolution along a different axis. In this work, we generalize common diffusion-based inverse single-image problem solvers for multi-image super-resolution (MISR) MRI. We show that the DPS likelihood correction allows an exactly-separable gradient decomposition across independently acquired measurements, enabling MISR without constructing a joint operator, modifying the diffusion model, or increasing network function evaluations. We derive MISR versions of DPS, DMAP, DPPS, and diffusion-based PnP/ADMM, and demonstrate substantial gains over SISR across $4\times/8\times/16\times$ anisotropic degradations. Our results achieve state-of-the-art super-resolution of anisotropic MRI volumes and, critically, enable reconstruction of near-isotropic anatomy from routine 2D multi-slice acquisitions, which are otherwise highly degraded in orthogonal views. --- ### Numina-Lean-Agent: An Open and General Agentic Reasoning System for Formal Mathematics Authors: Junqi Liu, Zihao Zhou, Zekai Zhu, Marco Dos Santos, Weikun He, Jiawei Liu, Ran Wang, Yunzhou Xie, Junqiao Zhao, Qiufeng Wang, Lihong Zhi, Jia Li et al. Published: 2026-01-20 Categories: cs.AI arXiv: https://arxiv.org/abs/2601.14027v1 Agentic systems have recently become the dominant paradigm for formal theorem proving, achieving strong performance by coordinating multiple models and tools. However, existing approaches often rely on task-specific pipelines and trained formal provers, limiting their flexibility and reproducibility. In this paper, we propose the paradigm that directly uses a general coding agent as a formal math reasoner. This paradigm is motivated by (1) A general coding agent provides a natural interface for diverse reasoning tasks beyond proving, (2) Performance can be improved by simply replacing the underlying base model, without training, and (3) MCP enables flexible extension and autonomous calling of specialized tools, avoiding complex design. Based on this paradigm, we introduce Numina-Lean-Agent, which combines Claude Code with Numina-Lean-MCP to enable autonomous interaction with Lean, retrieval of relevant theorems, informal proving and auxiliary reasoning tools. Using Claude Opus 4.5 as the base model, Numina-Lean-Agent solves all problems in Putnam 2025 (12 / 12), matching the best closed-source system. Beyond benchmark evaluation, we further demonstrate its generality by interacting with mathematicians to successfully formalize the Brascamp-Lieb theorem. We release Numina-Lean-Agent and all solutions at https://github.com/project-numina/numina-lean-agent. --- ### Universal Approximation Theorem for Input-Connected Multilayer Perceptrons Authors: Vugar Ismailov Published: 2026-01-20 Categories: cs.LG, cs.NE, math.FA arXiv: https://arxiv.org/abs/2601.14026v2 We present the Input-Connected Multilayer Perceptron (IC-MLP), a feedforward neural network architecture in which each hidden neuron receives, in addition to the outputs of the preceding layer, a direct affine connection from the raw input. We first study this architecture in the univariate setting and give an explicit and systematic description of IC-MLPs with an arbitrary finite number of hidden layers, including iterated formulas for the network functions. In this setting, we prove a universal approximation theorem showing that deep IC-MLPs can approximate any continuous function on a closed interval of the real line if and only if the activation function is nonlinear. We then extend the analysis to vector-valued inputs and establish a corresponding universal approximation theorem for continuous functions on compact subsets of $\mathbb{R}^n$. --- ### Credible CO2 Comparisons: A Machine Learning Approach to Vehicle Powertrain Assessment Authors: Rodrigo Pereira David, Luciano Araujo Dourado Filho, Daniel Marques da Silva, João Alfredo Cal-Braz Published: 2026-01-20 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2601.14022v1 Decarbonizing road transport requires consistent and transparent methods for comparing CO2 emissions across vehicle technologies. This paper proposes a machine learning-based framework for like-for-like operational assessment of internal combustion engine vehicles (ICEVs) and electric vehicles (EVs) under identical, real-world driving conditions. The approach isolates technology-specific effects by holding the observed speed profile and environmental context fixed, enabling direct comparison of powertrain performance. Recurrent neural network models are trained independently for each domain to learn the mapping from contextual driving variables (speed, acceleration, temperature) to internal actuation variables (torque, throttle) and instantaneous CO2-equivalent emission rates. This structure allows the construction of counterfactual scenarios that answer: What emissions would an EV have generated if it had followed the same driving profile as an ICEV? By aligning both vehicle types on a unified instantaneous emissions metric, the framework enables fair and reproducible evaluation of powertrain technologies. It offers a scalable foundation for credible, data-driven assessments of vehicle carbon performance under real-world operating conditions. --- ### MATE: Matryoshka Audio-Text Embeddings for Open-Vocabulary Keyword Spotting Authors: Youngmoon Jung, Myunghun Jung, Joon-Young Yang, Yong-Hyeok Lee, Jaeyoung Roh, Hoon-Young Cho Published: 2026-01-20 Categories: eess.AS, cs.AI arXiv: https://arxiv.org/abs/2601.14012v1 Open-vocabulary keyword spotting (KWS) with text-based enrollment has emerged as a flexible alternative to fixed-phrase triggers. Prior utterance-level matching methods, from an embedding-learning standpoint, learn embeddings at a single fixed dimensionality. We depart from this design and propose Matryoshka Audio-Text Embeddings (MATE), a dual-encoder framework that encodes multiple embedding granularities within a single vector via nested sub-embeddings ("prefixes"). Specifically, we introduce a PCA-guided prefix alignment: PCA-compressed versions of the full text embedding for each prefix size serve as teacher targets to align both audio and text prefixes. This alignment concentrates salient keyword cues in lower-dimensional prefixes, while higher dimensions add detail. MATE is trained with standard deep metric learning objectives for audio-text KWS, and is loss-agnostic. To our knowledge, this is the first application of matryoshka-style embeddings to KWS, achieving state-of-the-art results on WSJ and LibriPhrase without any inference overhead. --- ### BACH-V: Bridging Abstract and Concrete Human-Values in Large Language Models Authors: Junyu Zhang, Yipeng Kang, Jiong Guo, Jiayu Zhan, Junqi Wang Published: 2026-01-20 Categories: cs.CL arXiv: https://arxiv.org/abs/2601.14007v1 Do large language models (LLMs) genuinely understand abstract concepts, or merely manipulate them as statistical patterns? We introduce an abstraction-grounding framework that decomposes conceptual understanding into three capacities: interpretation of abstract concepts (Abstract-Abstract, A-A), grounding of abstractions in concrete events (Abstract-Concrete, A-C), and application of abstract principles to regulate concrete decisions (Concrete-Concrete, C-C). Using human values as a testbed - given their semantic richness and centrality to alignment - we employ probing (detecting value traces in internal activations) and steering (modifying representations to shift behavior). Across six open-source LLMs and ten value dimensions, probing shows that diagnostic probes trained solely on abstract value descriptions reliably detect the same values in concrete event narratives and decision reasoning, demonstrating cross-level transfer. Steering reveals an asymmetry: intervening on value representations causally shifts concrete judgments and decisions (A-C, C-C), yet leaves abstract interpretations unchanged (A-A), suggesting that encoded abstract values function as stable anchors rather than malleable activations. These findings indicate LLMs maintain structured value representations that bridge abstraction and action, providing a mechanistic and operational foundation for building value-driven autonomous AI systems with more transparent, generalizable alignment and control. --- ### Locate, Steer, and Improve: A Practical Survey of Actionable Mechanistic Interpretability in Large Language Models Authors: Hengyuan Zhang, Zhihao Zhang, Mingyang Wang, Zunhai Su, Yiwei Wang, Qianli Wang, Shuzhou Yuan, Ercong Nie, Xufeng Duan, Feijiang Han, Qibo Xue, Zeping Yu et al. Published: 2026-01-20 Categories: cs.CL arXiv: https://arxiv.org/abs/2601.14004v4 Mechanistic Interpretability (MI) has emerged as a vital approach to demystify the opaque decision-making of Large Language Models (LLMs). However, existing reviews primarily treat MI as an observational science, summarizing analytical insights while lacking a systematic framework for actionable intervention. To bridge this gap, we present a practical survey structured around the pipeline: "Locate, Steer, and Improve." We formally categorize Localizing (diagnosis) and Steering (intervention) methods based on specific Interpretable Objects to establish a rigorous intervention protocol. Furthermore, we demonstrate how this framework enables tangible improvements in Alignment, Capability, and Efficiency, effectively operationalizing MI as an actionable methodology for model optimization. The curated paper list of this work is available at https://github.com/rattlesnakey/Awesome-Actionable-MI-Survey. --- ### Cross-Sensory Brain Passage Retrieval: Scaling Beyond Visual to Audio Authors: Niall McGuire, Yashar Moshfeghi Published: 2026-01-20 Categories: cs.IR, cs.LG arXiv: https://arxiv.org/abs/2601.14001v2 Query formulation from internal information needs remains fundamentally challenging across all Information Retrieval paradigms due to cognitive complexity and physical impairments. Brain Passage Retrieval (BPR) addresses this by directly mapping EEG signals to passage representations without intermediate text translation. However, existing BPR research exclusively uses visual stimuli, leaving critical questions unanswered: Can auditory EEG enable effective retrieval for voice-based interfaces and visually impaired users? Can training on combined EEG datasets from different sensory modalities improve performance despite severe data scarcity? We present the first systematic investigation of auditory EEG for BPR and evaluate cross-sensory training benefits. Using dual encoder architectures with four pooling strategies (CLS, mean, max, multi-vector), we conduct controlled experiments comparing auditory-only, visual-only, and combined training on the Alice (auditory) and Nieuwland (visual) datasets. Results demonstrate that auditory EEG consistently outperforms visual EEG, and cross-sensory training with CLS pooling achieves substantial improvements over individual training: 31% in MRR (0.474), 43% in Hit@1 (0.314), and 28% in Hit@10 (0.858). Critically, combined auditory EEG models surpass BM25 text baselines (MRR: 0.474 vs 0.428), establishing neural queries as competitive with traditional retrieval whilst enabling accessible interfaces. These findings validate auditory neural interfaces for IR tasks and demonstrate that cross-sensory training addresses data scarcity whilst outperforming single-modality approaches Code: https://github.com/NiallMcguire/Audio_BPR --- ### Group-Invariant Unsupervised Skill Discovery: Symmetry-aware Skill Representations for Generalizable Behavior Authors: Junwoo Chang, Joseph Park, Roberto Horowitz, Jongmin Lee, Jongeun Choi Published: 2026-01-20 Categories: cs.RO, cs.LG arXiv: https://arxiv.org/abs/2601.14000v1 Unsupervised skill discovery aims to acquire behavior primitives that improve exploration and accelerate downstream task learning. However, existing approaches often ignore the geometric symmetries of physical environments, leading to redundant behaviors and sample inefficiency. To address this, we introduce Group-Invariant Skill Discovery (GISD), a framework that explicitly embeds group structure into the skill discovery objective. Our approach is grounded in a theoretical guarantee: we prove that in group-symmetric environments, the standard Wasserstein dependency measure admits a globally optimal solution comprised of an equivariant policy and a group-invariant scoring function. Motivated by this, we formulate the Group-Invariant Wasserstein dependency measure, which restricts the optimization to this symmetry-aware subspace without loss of optimality. Practically, we parameterize the scoring function using a group Fourier representation and define the intrinsic reward via the alignment of equivariant latent features, ensuring that the discovered skills generalize systematically under group transformations. Experiments on state-based and pixel-based locomotion benchmarks demonstrate that GISD achieves broader state-space coverage and improved efficiency in downstream task learning compared to a strong baseline. --- ### DAME: Duration-Aware Matryoshka Embedding for Duration-Robust Speaker Verification Authors: Youngmoon Jung, Joon-Young Yang, Ju-ho Kim, Jaeyoung Roh, Chang Woo Han, Hoon-Young Cho Published: 2026-01-20 Categories: eess.AS, cs.AI arXiv: https://arxiv.org/abs/2601.13999v1 Short-utterance speaker verification remains challenging due to limited speaker-discriminative cues in short speech segments. While existing methods focus on enhancing speaker encoders, the embedding learning strategy still forces a single fixed-dimensional representation reused for utterances of any length, leaving capacity misaligned with the information available at different durations. We propose Duration-Aware Matryoshka Embedding (DAME), a model-agnostic framework that builds a nested hierarchy of sub-embeddings aligned to utterance durations: lower-dimensional representations capture compact speaker traits from short utterances, while higher dimensions encode richer details from longer speech. DAME supports both training from scratch and fine-tuning, and serves as a direct alternative to conventional large-margin fine-tuning, consistently improving performance across durations. On the VoxCeleb1-O/E/H and VOiCES evaluation sets, DAME consistently reduces the equal error rate on 1-s and other short-duration trials, while maintaining full-length performance with no additional inference cost. These gains generalize across various speaker encoder architectures under both general training and fine-tuning setups. --- ### From Tags to Trees: Structuring Fine-Grained Knowledge for Controllable Data Selection in LLM Instruction Tuning Authors: Zihan Niu, Wenping Hu, Junmin Chen, Xiyue Wang, Tong Xu, Ruiming Tang Published: 2026-01-20 Categories: cs.CL arXiv: https://arxiv.org/abs/2601.13995v1 Effective and controllable data selection is critical for LLM instruction tuning, especially with massive open-source datasets. Existing approaches primarily rely on instance-level quality scores, or diversity metrics based on embedding clusters or semantic tags. However, constrained by the flatness of embedding spaces or the coarseness of tags, these approaches overlook fine-grained knowledge and its intrinsic hierarchical dependencies, consequently hindering precise data valuation and knowledge-aligned sampling. To address this challenge, we propose Tree-aware Aligned Global Sampling (TAGS), a unified framework that leverages a knowledge tree built from fine-grained tags, thereby enabling joint control of global quality, diversity, and target alignment. Using an LLM-based tagger, we extract atomic knowledge concepts, which are organized into a global tree through bottom-up hierarchical clustering. By grounding data instances onto this tree, a tree-aware metric then quantifies data quality and diversity, facilitating effective sampling. Our controllable sampling strategy maximizes tree-level information gain and enforces leaf-level alignment via KL-divergence for specific domains. Extensive experiments demonstrate that TAGS significantly outperforms state-of-the-art baselines. Notably, it surpasses the full-dataset model by \textbf{+5.84\%} using only \textbf{5\%} of the data, while our aligned sampling strategy further boosts average performance by \textbf{+4.24\%}. --- ### torch-sla: Differentiable Sparse Linear Algebra with Adjoint Solvers and Sparse Tensor Parallelism for PyTorch Authors: Mingyuan Chi, Shizheng Wen Published: 2026-01-20 Categories: cs.DC, cs.AI arXiv: https://arxiv.org/abs/2601.13994v3 Differentiable sparse linear algebra is foundational for scientific machine learning, yet PyTorch lacks a unified library for it: torch.sparse provides only low-level kernels and a non-differentiable, CPU-only spsolve, and torch.linalg is dense-only. We present torch-sla, an open-source library that fills this gap. It exposes a single autograd-aware API for direct, iterative, nonlinear, and eigenvalue solvers across five interchangeable backends -- SciPy and Eigen on CPU, cuDSS, CuPy, and a PyTorch-native iterative solver on GPU -- with automatic dispatch by device and problem size. The library further supports batched solves over shared or distinct sparsity patterns and distributed multi-GPU execution via domain decomposition with halo exchange. These capabilities are made scalable by an O(1)-graph adjoint differentiation framework and an autograd-compatible distributed halo-exchange layer. The library is available at https://www.torchsla.com/. --- ### "The Whole Is Greater Than the Sum of Its Parts": A Compatibility-Aware Multi-Teacher CoT Distillation Framework Authors: Jin Cui, Jiaqi Guo, Ruixuan Yang, Jiayi Lu, Jiepeng Zhou, Jiajun Xu, Jiangcheng Song, Boran Zhao, Pengju Ren Published: 2026-01-20 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2601.13992v2 Chain-of-Thought (CoT) reasoning empowers Large Language Models (LLMs) with remarkable capabilities but typically requires prohibitive parameter scales. CoT distillation has emerged as a promising paradigm to transfer reasoning prowess into compact Student Models (SLMs), but existing approaches often rely on a solitary teacher, capping the student's potential since individual LLMs often exhibit distinct capability biases and may suffer from catastrophic forgetting. While leveraging diverse teachers seems appealing, effectively fusing their supervisions remains challenging: teacher-student incompatibility risks amplifying hallucinations, and passive supervision fails to ensure genuine logic internalization. To address this, we introduce COMPACT, a framework that adaptively fuses supervisions from different teachers by dynamically weighting teacher gradients based on the student's real-time compatibility evaluated by a multi-dimensional metric: (1) Graph-based Consensus to filter misleading rationales by identifying mainstream reasoning paths; (2) Mutual-Information-based Adaptability to detect "epiphany moments" for genuinely understanding the reasoning process rather than merely imitating; and (3) Loss-based Difficulty to assess student receptivity to the teacher's guidance and prevent negative transfer. Extensive experiments and latent space analysis demonstrate that COMPACT effectively integrates diverse reasoning capabilities without damaging the model's original knowledge structure, achieving state-of-the-art performance on various benchmarks while mitigating catastrophic forgetting. --- ### Linearized subspace refinement framework to expose hidden accuracy in trained neural networks Authors: Wenbo Cao, Weiwei Zhang Published: 2026-01-20 Categories: cs.LG arXiv: https://arxiv.org/abs/2601.13989v2 Neural networks trained by gradient-based methods often exhibit optimization-induced accuracy plateaus in scientific machine learning tasks. We present Linearized Subspace Refinement (LSR), an architecture-agnostic post-training framework that exploits the local linearized model at a fixed trained state. By solving a reduced direct least-squares problem in a Jacobian-defined low-dimensional space, LSR computes a subspace-optimal linearized correction and yields a refined predictor with markedly improved accuracy. Across function approximation, data-driven operator learning, physics-informed operator fine-tuning, and noisy inverse problems, LSR shows that standard nonlinear training can remain far above this subspace-attainable error level. Similar accuracy plateaus persist even for the convex quadratic problem from local linearization when solved with standard iterative optimizers, identifying numerical ill-conditioning as a primary bottleneck. LSR frequently delivers order-of-magnitude error reductions, while the subspace rank provides an explicit capacity-control mechanism that balances correction strength, numerical stability, and noise sensitivity. Together, LSR exposes conditioning-limited attainable accuracy in trained-state linearized models and provides direct access to it. --- ### SHARE: A Fully Unsupervised Framework for Single Hyperspectral Image Restoration Authors: Jiangwei Xie, Zhang Wen, Mike Davies, Dongdong Chen Published: 2026-01-20 Categories: eess.IV, cs.CV arXiv: https://arxiv.org/abs/2601.13987v1 Hyperspectral image (HSI) restoration is a fundamental challenge in computational imaging and computer vision. It involves ill-posed inverse problems, such as inpainting and super-resolution. Although deep learning methods have transformed the field through data-driven learning, their effectiveness hinges on access to meticulously curated ground-truth datasets. This fundamentally restricts their applicability in real-world scenarios where such data is unavailable. This paper presents SHARE (Single Hyperspectral Image Restoration with Equivariance), a fully unsupervised framework that unifies geometric equivariance principles with low-rank spectral modelling to eliminate the need for ground truth. SHARE's core concept is to exploit the intrinsic invariance of hyperspectral structures under differentiable geometric transformations (e.g. rotations and scaling) to derive self-supervision signals through equivariance consistency constraints. Our novel Dynamic Adaptive Spectral Attention (DASA) module further enhances this paradigm shift by explicitly encoding the global low-rank property of HSI and adaptively refining local spectral-spatial correlations through learnable attention mechanisms. Extensive experiments on HSI inpainting and super-resolution tasks demonstrate the effectiveness of SHARE. Our method outperforms many state-of-the-art unsupervised approaches and achieves performance comparable to that of supervised methods. We hope that our approach will shed new light on HSI restoration and broader scientific imaging scenarios. The code will be released at https://github.com/xuwayyy/SHARE. --- ### Equivariant Learning for Unsupervised Image Dehazing Authors: Zhang Wen, Jiangwei Xie, Dongdong Chen Published: 2026-01-20 Categories: cs.CV, eess.IV arXiv: https://arxiv.org/abs/2601.13986v1 Image Dehazing (ID) aims to produce a clear image from an observation contaminated by haze. Current ID methods typically rely on carefully crafted priors or extensive haze-free ground truth, both of which are expensive or impractical to acquire, particularly in the context of scientific imaging. We propose a new unsupervised learning framework called Equivariant Image Dehazing (EID) that exploits the symmetry of image signals to restore clarity to hazy observations. By enforcing haze consistency and systematic equivariance, EID can recover clear patterns directly from raw, hazy images. Additionally, we propose an adversarial learning strategy to model unknown haze physics and facilitate EID learning. Experiments on two scientific image dehazing benchmarks (including cell microscopy and medical endoscopy) and on natural image dehazing have demonstrated that EID significantly outperforms state-of-the-art approaches. By unifying equivariant learning with modelling haze physics, we hope that EID will enable more versatile and effective haze removal in scientific imaging. Code and datasets will be published. --- ### FantasyVLN: Unified Multimodal Chain-of-Thought Reasoning for Vision-Language Navigation Authors: Jing Zuo, Lingzhou Mu, Fan Jiang, Chengcheng Ma, Mu Xu, Yonggang Qi Published: 2026-01-20 Categories: cs.CV, cs.RO arXiv: https://arxiv.org/abs/2601.13976v2 Achieving human-level performance in Vision-and-Language Navigation (VLN) requires an embodied agent to jointly understand multimodal instructions and visual-spatial context while reasoning over long action sequences. Recent works, such as NavCoT and NavGPT-2, demonstrate the potential of Chain-of-Thought (CoT) reasoning for improving interpretability and long-horizon planning. Moreover, multimodal extensions like OctoNav-R1 and CoT-VLA further validate CoT as a promising pathway toward human-like navigation reasoning. However, existing approaches face critical drawbacks: purely textual CoTs lack spatial grounding and easily overfit to sparse annotated reasoning steps, while multimodal CoTs incur severe token inflation by generating imagined visual observations, making real-time navigation impractical. In this work, we propose FantasyVLN, a unified implicit reasoning framework that preserves the benefits of CoT reasoning without explicit token overhead. Specifically, imagined visual tokens are encoded into a compact latent space using a pretrained Visual AutoRegressor (VAR) during CoT reasoning training, and the model jointly learns from textual, visual, and multimodal CoT modes under a unified multi-CoT strategy. At inference, our model performs direct instruction-to-action mapping while still enjoying reasoning-aware representations. Extensive experiments on LH-VLN show that our approach achieves reasoning-aware yet real-time navigation, improving success rates and efficiency while reducing inference latency by an order of magnitude compared to explicit CoT methods. --- ### Harmonizing the Deep: A Unified Information Pipeline for Robust Marine Biodiversity Assessment Across Heterogeneous Domains Authors: Marco Piccolo, Qiwei Han, Astrid van Toor, Joachim Vanneste Published: 2026-01-20 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2601.13975v1 Marine biodiversity monitoring requires scalability and reliability across complex underwater environments to support conservation and invasive-species management. Yet existing detection solutions often exhibit a pronounced deployment gap, with performance degrading sharply when transferred to new sites. This work establishes the foundational detection layer for a multi-year invasive species monitoring initiative targeting Arctic and Atlantic marine ecosystems. We address this challenge by developing a Unified Information Pipeline that standardises heterogeneous datasets into a comparable information flow and evaluates a fixed, deployment-relevant detector under controlled cross-domain protocols. Across multiple domains, we find that structural factors, such as scene composition, object density, and contextual redundancy, explain cross-domain performance loss more strongly than visual degradation such as turbidity, with sparse scenes inducing a characteristic "Context Collapse" failure mode. We further validate operational feasibility by benchmarking inference on low-cost edge hardware, showing that runtime optimisation enables practical sampling rates for remote monitoring. The results shift emphasis from image enhancement toward structure-aware reliability, providing a democratised tool for consistent marine ecosystem assessment. --- ### STEC: A Reference-Free Spatio-Temporal Entropy Coverage Metric for Evaluating Sampled Video Frames Authors: Shih-Yao Lin Published: 2026-01-20 Categories: cs.CV arXiv: https://arxiv.org/abs/2601.13974v1 Frame sampling is a fundamental component in video understanding and video--language model pipelines, yet evaluating the quality of sampled frames remains challenging. Existing evaluation metrics primarily focus on perceptual quality or reconstruction fidelity, and are not designed to assess whether a set of sampled frames adequately captures informative and representative video content. We propose Spatio-Temporal Entropy Coverage (STEC), a simple and non-reference metric for evaluating the effectiveness of video frame sampling. STEC builds upon Spatio-Temporal Frame Entropy (STFE), which measures per-frame spatial information via entropy-based structural complexity, and evaluates sampled frames based on their temporal coverage and redundancy. By jointly modeling spatial information strength, temporal dispersion, and non-redundancy, STEC provides a principled and lightweight measure of sampling quality. Experiments on the MSR-VTT test-1k benchmark demonstrate that STEC clearly differentiates common sampling strategies, including random, uniform, and content-aware methods. We further show that STEC reveals robustness patterns across individual videos that are not captured by average performance alone, highlighting its practical value as a general-purpose evaluation tool for efficient video understanding. We emphasize that STEC is not designed to predict downstream task accuracy, but to provide a task-agnostic diagnostic signal for analyzing frame sampling behavior under constrained budgets. --- ### Autonomous Knowledge Graph Exploration with Adaptive Breadth-Depth Retrieval Authors: Joaquín Polonuer, Lucas Vittor, Iñaki Arango, Ayush Noori, David A. Clifton, Luciano Del Corro, Marinka Zitnik Published: 2026-01-20 Categories: cs.AI, cs.IR, cs.LG arXiv: https://arxiv.org/abs/2601.13969v2 Retrieving evidence for language model queries from knowledge graphs requires balancing broad search across the graph with multi-hop traversal to follow relational links. Similarity-based retrievers provide coverage but remain shallow, whereas traversal-based methods rely on selecting seed nodes to start exploration, which can fail when queries span multiple entities and relations. We introduce ARK: Adaptive Retriever of Knowledge, a tool-using KG retriever that gives a language model control over this breadth-depth tradeoff using a two-operation toolset: global lexical search over node descriptors and one-hop neighborhood exploration that composes into multi-hop traversal. ARK alternates between breadth-oriented discovery and depth-oriented expansion without depending on a fragile seed selection, a pre-set hop depth, or requiring retrieval training. ARK adapts tool use to queries, using global search for language-heavy queries and neighborhood exploration for relation-heavy queries. On STaRK, ARK reaches 59.1% average Hit@1 and 67.4 average MRR, improving average Hit@1 by up to 31.4% and average MRR by up to 28.0% over retrieval-based and agent-based training-free methods. Finally, we distill ARK's tool-use trajectories from a large teacher into an 8B model via label-free imitation, improving Hit@1 by +7.0, +26.6, and +13.5 absolute points over the base 8B model on AMAZON, MAG, and PRIME datasets, respectively, while retaining up to 98.5% of the teacher's Hit@1 rate. --- ### RL-BioAug: Label-Efficient Reinforcement Learning for Self-Supervised EEG Representation Learning Authors: Cheol-Hui Lee, Hwa-Yeon Lee, Dong-Joo Kim Published: 2026-01-20 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2601.13964v2 The quality of data augmentation serves as a critical determinant for the performance of contrastive learning in EEG tasks. Although this paradigm is promising for utilizing unlabeled data, static or random augmentation strategies often fail to preserve intrinsic information due to the non-stationarity of EEG signals where statistical properties change over time. To address this, we propose RL-BioAug, a framework that leverages a label-efficient reinforcement learning (RL) agent to autonomously determine optimal augmentation policies. While utilizing only a minimal fraction (10%) of labeled data to guide the agent's policy, our method enables the encoder to learn robust representations in a strictly self-supervised manner. Experimental results demonstrate that RL-BioAug significantly outperforms the random selection strategy, achieving substantial improvements of 9.69% and 8.80% in Macro-F1 score on the Sleep-EDFX and CHB-MIT datasets, respectively. Notably, this agent mainly chose optimal strategies for each task--for example, Time Masking with a 62% probability for sleep stage classification and Crop & Resize with a 77% probability for seizure detection. Our framework suggests its potential to replace conventional heuristic-based augmentations and establish a new autonomous paradigm for data augmentation. The source code is available at https://github.com/dlcjfgmlnasa/RL-BioAug. --- ### DExTeR: Weakly Semi-Supervised Object Detection with Class and Instance Experts for Medical Imaging Authors: Adrien Meyer, Didier Mutter, Nicolas Padoy Published: 2026-01-20 Categories: cs.CV arXiv: https://arxiv.org/abs/2601.13954v1 Detecting anatomical landmarks in medical imaging is essential for diagnosis and intervention guidance. However, object detection models rely on costly bounding box annotations, limiting scalability. Weakly Semi-Supervised Object Detection (WSSOD) with point annotations proposes annotating each instance with a single point, minimizing annotation time while preserving localization signals. A Point-to-Box teacher model, trained on a small box-labeled subset, converts these point annotations into pseudo-box labels to train a student detector. Yet, medical imagery presents unique challenges, including overlapping anatomy, variable object sizes, and elusive structures, which hinder accurate bounding box inference. To overcome these challenges, we introduce DExTeR (DETR with Experts), a transformer-based Point-to-Box regressor tailored for medical imaging. Built upon Point-DETR, DExTeR encodes single-point annotations as object queries, refining feature extraction with the proposed class-guided deformable attention, which guides attention sampling using point coordinates and class labels to capture class-specific characteristics. To improve discrimination in complex structures, it introduces CLICK-MoE (CLass, Instance, and Common Knowledge Mixture of Experts), decoupling class and instance representations to reduce confusion among adjacent or overlapping instances. Finally, we implement a multi-point training strategy which promotes prediction consistency across different point placements, improving robustness to annotation variability. DExTeR achieves state-of-the-art performance across three datasets spanning different medical domains (endoscopy, chest X-rays, and endoscopic ultrasound) highlighting its potential to reduce annotation costs while maintaining high detection accuracy. --- ### Differentiable Logic Synthesis: Spectral Coefficient Selection via Sinkhorn-Constrained Composition Authors: Gorgi Pavlov Published: 2026-01-20 Categories: cs.LG, cs.AR, cs.LO arXiv: https://arxiv.org/abs/2601.13953v3 Learning precise Boolean logic via gradient descent remains challenging: neural networks typically converge to "fuzzy" approximations that degrade under quantization. We introduce Hierarchical Spectral Composition, a differentiable architecture that selects spectral coefficients from a frozen Boolean Fourier basis and composes them via Sinkhorn-constrained routing with column-sign modulation. Our approach draws on recent insights from Manifold-Constrained Hyper-Connections (mHC), which demonstrated that projecting routing matrices onto the Birkhoff polytope preserves identity mappings and stabilizes large-scale training. We adapt this framework to logic synthesis, adding column-sign modulation to enable Boolean negation -- a capability absent in standard doubly stochastic routing. We validate our approach across four phases of increasing complexity: (1) For n=2 (16 Boolean operations over 4-dim basis), gradient descent achieves 100% accuracy with zero routing drift and zero-loss quantization to ternary masks. (2) For n=3 (10 three-variable operations), gradient descent achieves 76% accuracy, but exhaustive enumeration over 3^8 = 6561 configurations proves that optimal ternary masks exist for all operations (100% accuracy, 39% sparsity). (3) For n=4 (10 four-variable operations over 16-dim basis), spectral synthesis -- combining exact Walsh-Hadamard coefficients, ternary quantization, and MCMC refinement with parallel tempering -- achieves 100% accuracy on all operations. This progression establishes (a) that ternary polynomial threshold representations exist for all tested functions, and (b) that finding them requires methods beyond pure gradient descent as dimensionality grows. All operations enable single-cycle combinational logic inference at 10,959 MOps/s on GPU, demonstrating viability for hardware-efficient neuro-symbolic logic synthesis. --- ### VTONGuard: Automatic Detection and Authentication of AI-Generated Virtual Try-On Content Authors: Shengyi Wu, Yan Hong, Shengyao Chen, Zheng Wang, Xianbing Sun, Jiahui Zhan, Jun Lan, Jianfu Zhang Published: 2026-01-20 Categories: cs.CV arXiv: https://arxiv.org/abs/2601.13951v1 With the rapid advancement of generative AI, virtual try-on (VTON) systems are becoming increasingly common in e-commerce and digital entertainment. However, the growing realism of AI-generated try-on content raises pressing concerns about authenticity and responsible use. To address this, we present VTONGuard, a large-scale benchmark dataset containing over 775,000 real and synthetic try-on images. The dataset covers diverse real-world conditions, including variations in pose, background, and garment styles, and provides both authentic and manipulated examples. Based on this benchmark, we conduct a systematic evaluation of multiple detection paradigms under unified training and testing protocols. Our results reveal each method's strengths and weaknesses and highlight the persistent challenge of cross-paradigm generalization. To further advance detection, we design a multi-task framework that integrates auxiliary segmentation to enhance boundary-aware feature learning, achieving the best overall performance on VTONGuard. We expect this benchmark to enable fair comparisons, facilitate the development of more robust detection models, and promote the safe and responsible deployment of VTON technologies in practice. --- ### Stream-Voice-Anon: Enhancing Utility of Real-Time Speaker Anonymization via Neural Audio Codec and Language Models Authors: Nikita Kuzmin, Songting Liu, Kong Aik Lee, Eng Siong Chng Published: 2026-01-20 Categories: eess.AS, cs.AI arXiv: https://arxiv.org/abs/2601.13948v3 Protecting speaker identity is crucial for online voice applications, yet streaming speaker anonymization (SA) remains underexplored. Recent research has demonstrated that neural audio codec (NAC) provides superior speaker feature disentanglement and linguistic fidelity. NAC can also be used with causal language models (LM) to enhance linguistic fidelity and prompt control for streaming tasks. However, existing NAC-based online LM systems are designed for voice conversion (VC) rather than anonymization, lacking the techniques required for privacy protection. Building on these advances, we present Stream-Voice-Anon, which adapts modern causal LM-based NAC architectures specifically for streaming SA by integrating anonymization techniques. Our anonymization approach incorporates pseudo-speaker representation sampling, a speaker embedding mixing and diverse prompt selection strategies for LM conditioning that leverage the disentanglement properties of quantized content codes to prevent speaker information leakage. Additionally, we compare dynamic and fixed delay configurations to explore latency-privacy trade-offs in real-time scenarios. Under the VoicePrivacy 2024 Challenge protocol, Stream-Voice-Anon achieves substantial improvements in intelligibility (up to 46% relative WER reduction) and emotion preservation (up to 28% UAR relative) compared to the previous state-of-the-art streaming method DarkStream while maintaining comparable latency (180ms vs 200ms) and privacy protection against lazy-informed attackers, though showing 15% relative degradation against semi-informed attackers. --- ### Efficient Coordination with the System-Level Shared State: An Embodied-AI Native Modular Framework Authors: Yixuan Deng, Tongrun Wu, Donghao Wu, Zeyu Wei, Jiayuan Wang, Zhenglong Sun, Yuqing Tang, Xiaoqiang Ji Published: 2026-01-20 Categories: cs.RO, cs.LG arXiv: https://arxiv.org/abs/2601.13945v1 As Embodied AI systems move from research prototypes to real world deployments, they tend to evolve rapidly while remaining reliable under workload changes and partial failures. In practice, many deployments are only partially decoupled: middleware moves messages, but shared context and feedback semantics are implicit, causing interface drift, cross-module interference, and brittle recovery at scale. We present ANCHOR, a modular framework that makes decoupling and robustness explicit system-level primitives. ANCHOR separates (i) Canonical Records, an evolvable contract for the standardized shared state, from (ii) a communication bus for many-to-many dissemination and feedback-oriented coordination, forming an inspectable end-to-end loop. We validate closed-loop feasibility on a de-identified workflow instantiation, characterize latency distributions under varying payload sizes and publish rates, and demonstrate automatic stream resumption after hard crashes and restarts even with shared-memory loss. Overall, ANCHOR turns ad-hoc integration glue into explicit contracts, enabling controlled degradation under load and self-healing recovery for scalable deployment of closed-loop AI systems. --- ### Glance-or-Gaze: Incentivizing LMMs to Adaptively Focus Search via Reinforcement Learning Authors: Hongbo Bai, Yujin Zhou, Yile Wu, Chi-Min Chan, Pengcheng Wen, Kunhao Pan, Sirui Han, Yike Guo Published: 2026-01-20 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2601.13942v2 Large Multimodal Models (LMMs) have achieved remarkable success in visual understanding, yet they struggle with knowledge-intensive queries involving long-tail entities or evolving information due to static parametric knowledge. Recent search-augmented approaches attempt to address this limitation, but existing methods rely on indiscriminate whole-image retrieval that introduces substantial visual redundancy and noise, and lack deep iterative reflection, limiting their effectiveness on complex visual queries. To overcome these challenges, we propose Glance-or-Gaze (GoG), a fully autonomous framework that shifts from passive perception to active visual planning. GoG introduces a Selective Gaze mechanism that dynamically chooses whether to glance at global context or gaze into high-value regions, filtering irrelevant information before retrieval. We design a dual-stage training strategy: Reflective GoG Behavior Alignment via supervised fine-tuning instills the fundamental GoG paradigm, while Complexity-Adaptive Reinforcement Learning further enhances the model's capability to handle complex queries through iterative reasoning. Experiments across six benchmarks demonstrate state-of-the-art performance. Ablation studies confirm that both Selective Gaze and complexity-adaptive RL are essential for effective visual search. --- ### IF-GEO: Conflict-Aware Instruction Fusion for Multi-Query Generative Engine Optimization Authors: Heyang Zhou, JiaJia Chen, Xiaolu Chen, Jie Bao, Zhen Chen, Yong Liao Published: 2026-01-20 Categories: cs.IR, cs.AI arXiv: https://arxiv.org/abs/2601.13938v1 As Generative Engines revolutionize information retrieval by synthesizing direct answers from retrieved sources, ensuring source visibility becomes a significant challenge. Improving it through targeted content revisions is a practical strategy termed Generative Engine Optimization (GEO). However, optimizing a document for diverse queries presents a constrained optimization challenge where heterogeneous queries often impose conflicting and competing revision requirements under a limited content budget. To address this challenge, we propose IF-GEO, a "diverge-then-converge" framework comprising two phases: (i) mining distinct optimization preferences from representative latent queries; (ii) synthesizing a Global Revision Blueprint for guided editing by coordinating preferences via conflict-aware instruction fusion. To explicitly quantify IF-GEO's objective of cross-query stability, we introduce risk-aware stability metrics. Experiments on multi-query benchmarks demonstrate that IF-GEO achieves substantial performance gains while maintaining robustness across diverse retrieval scenarios. ---