Beyond Scaling Law: A Data-Efficient Distillation Framework for Reasoning

Large language models (LLMs) demonstrate remarkable reasoning capabilities in tasks such as algorithmic coding and mathematical problem-solving. Recent methods have improved reasoning through expanded corpus and multistage training combining reinforcement learning and supervised fine-tuning. Although some methods suggest that small but targeted dataset can incentivize reasoning via only distillation, a reasoning scaling laws is still taking shape, increasing computational costs. To address this, we propose a data-efficient distillation framework (DED) that optimizes the Pareto frontier of reasoning distillation. Inspired by the on-policy learning and diverse roll-out strategies of reinforcement learning, the key idea of our approach is threefold: (1) We identify that benchmark scores alone do not determine an effective teacher model. Through comprehensive comparisons of leading reasoning LLMs, we develop a method to select an optimal teacher model. (2) While scaling distillation can enhance reasoning, it often degrades out-of-domain performance. A carefully curated, smaller corpus achieves a balanced trade-off between in-domain and out-of-domain capabilities. (3) Diverse reasoning trajectories encourage the student model to develop robust reasoning skills. We validate our method through evaluations on mathematical reasoning (AIME 2024/2025, MATH-500) and code generation (LiveCodeBench), achieving state-of-the-art results with only 0.8k carefully curated examples, bypassing the need for extensive scaling. Our systematic analysis demonstrates that DED outperforms existing methods by considering factors beyond superficial hardness, token length, or teacher model capability. This work offers a practical and efficient pathway to advanced reasoning while preserving general capabilities.

Hybrid Batch Normalisation: Resolving the Dilemma of Batch Normalisation in Federated Learning

Batch Normalisation (BN) is widely used in conventional deep neural network training to harmonise the input-output distributions for each batch of data. However, federated learning, a distributed learning paradigm, faces the challenge of dealing with non-independent and identically distributed data among the client nodes. Due to the lack of a coherent methodology for updating BN statistical parameters, standard BN degrades the federated learning performance. To this end, it is urgent to explore an alternative normalisation solution for federated learning. In this work, we resolve the dilemma of the BN layer in federated learning by developing a customised normalisation approach, Hybrid Batch Normalisation (HBN). HBN separates the update of statistical parameters (i.e. , means and variances used for evaluation) from that of learnable parameters (i.e. , parameters that require gradient updates), obtaining unbiased estimates of global statistical parameters in distributed scenarios. In contrast with the existing solutions, we emphasise the supportive power of global statistics for federated learning. The HBN layer introduces a learnable hybrid distribution factor, allowing each computing node to adaptively mix the statistical parameters of the current batch with the global statistics. Our HBN can serve as a powerful plugin to advance federated learning performance. It reflects promising merits across a wide range of federated learning settings, especially for small batch sizes and heterogeneous data.

Select2Reason: Efficient Instruction-Tuning Data Selection for Long-CoT Reasoning

A practical approach to activate long chain-of-thoughts reasoning ability in pre-trained large language models is to perform supervised fine-tuning on instruction datasets synthesized by strong Large Reasoning Models such as DeepSeek-R1, offering a cost-effective alternative to reinforcement learning. However, large-scale instruction sets with more than 100k samples incur significant training overhead, while effective strategies for automatic long-CoT instruction selection still remain unexplored. In this work, we propose Select2Reason, a novel and efficient instruction-tuning data selection framework for long-CoT reasoning. From the perspective of emergence of rethinking behaviors like self-correction and backtracking, we investigate common metrics that may determine the quality of long-CoT reasoning instructions. Select2Reason leverages a quantifier to estimate difficulty of question and jointly incorporates a reasoning trace length-based heuristic through a weighted scheme for ranking to prioritize high-utility examples. Empirical results on OpenR1-Math-220k demonstrate that fine-tuning LLM on only 10% of the data selected by Select2Reason achieves performance competitive with or superior to full-data tuning and open-source baseline OpenR1-Qwen-7B across three competition-level and six comprehensive mathematical benchmarks. Further experiments highlight the scalability in varying data size, efficiency during inference, and its adaptability to other instruction pools with minimal cost.

iPad: Iterative Proposal-centric End-to-End Autonomous Driving

End-to-end (E2E) autonomous driving systems offer a promising alternative to traditional modular pipelines by reducing information loss and error accumulation, with significant potential to enhance both mobility and safety. However, most existing E2E approaches directly generate plans based on dense bird's-eye view (BEV) grid features, leading to inefficiency and limited planning awareness. To address these limitations, we propose iterative Proposal-centric autonomous driving (iPad), a novel framework that places proposals - a set of candidate future plans - at the center of feature extraction and auxiliary tasks. Central to iPad is ProFormer, a BEV encoder that iteratively refines proposals and their associated features through proposal-anchored attention, effectively fusing multi-view image data. Additionally, we introduce two lightweight, proposal-centric auxiliary tasks - mapping and prediction - that improve planning quality with minimal computational overhead. Extensive experiments on the NAVSIM and CARLA Bench2Drive benchmarks demonstrate that iPad achieves state-of-the-art performance while being significantly more efficient than prior leading methods.

From Deep Learning to LLMs: A survey of AI in Quantitative Investment

Quantitative investment (quant) is an emerging, technology-driven approach in asset management, increasingy shaped by advancements in artificial intelligence. Recent advances in deep learning and large language models (LLMs) for quant finance have improved predictive modeling and enabled agent-based automation, suggesting a potential paradigm shift in this field. In this survey, taking alpha strategy as a representative example, we explore how AI contributes to the quantitative investment pipeline. We first examine the early stage of quant research, centered on human-crafted features and traditional statistical models with an established alpha pipeline. We then discuss the rise of deep learning, which enabled scalable modeling across the entire pipeline from data processing to order execution. Building on this, we highlight the emerging role of LLMs in extending AI beyond prediction, empowering autonomous agents to process unstructured data, generate alphas, and support self-iterative workflows.

MultiBARF: Integrating Imagery of Different Wavelength Regions by Using Neural Radiance Fields

Optical sensor applications have become popular through digital transformation. Linking observed data to real-world locations and combining different image sensors is essential to make the applications practical and efficient. However, data preparation to try different sensor combinations requires high sensing and image processing expertise. To make data preparation easier for users unfamiliar with sensing and image processing, we have developed MultiBARF. This method replaces the co-registration and geometric calibration by synthesizing pairs of two different sensor images and depth images at assigned viewpoints. Our method extends Bundle Adjusting Neural Radiance Fields(BARF), a deep neural network-based novel view synthesis method, for the two imagers. Through experiments on visible light and thermographic images, we demonstrate that our method superimposes two color channels of those sensor images on NeRF.

One Model for ALL: Low-Level Task Interaction Is a Key to Task-Agnostic Image Fusion

Advanced image fusion methods mostly prioritise high-level missions, where task interaction struggles with semantic gaps, requiring complex bridging mechanisms. In contrast, we propose to leverage low-level vision tasks from digital photography fusion, allowing for effective feature interaction through pixel-level supervision. This new paradigm provides strong guidance for unsupervised multimodal fusion without relying on abstract semantics, enhancing task-shared feature learning for broader applicability. Owning to the hybrid image features and enhanced universal representations, the proposed GIFNet supports diverse fusion tasks, achieving high performance across both seen and unseen scenarios with a single model. Uniquely, experimental results reveal that our framework also supports single-modality enhancement, offering superior flexibility for practical applications. Our code will be available at https://github.com/AWCXV/GIFNet.

Subpixel Edge Localization Based on Converted Intensity Summation under Stable Edge Region

To satisfy the rigorous requirements of precise edge detection in critical high-accuracy measurements, this article proposes a series of efficient approaches for localizing subpixel edge. In contrast to the fitting based methods, which consider pixel intensity as a sample value derived from a specific model. We take an innovative perspective by assuming that the intensity at the pixel level can be interpreted as a local integral mapping in the intensity model for subpixel localization. Consequently, we propose a straightforward subpixel edge localization method called Converted Intensity Summation (CIS). To address the limited robustness associated with focusing solely on the localization of individual edge points, a Stable Edge Region (SER) based algorithm is presented to alleviate local interference near edges. Given the observation that the consistency of edge statistics exists in the local region, the algorithm seeks correlated stable regions in the vicinity of edges to facilitate the acquisition of robust parameters and achieve higher precision positioning. In addition, an edge complement method based on extension-adjustment is also introduced to rectify the irregular edges through the efficient migration of SERs. A large number of experiments are conducted on both synthetic and real image datasets which cover common edge patterns as well as various real scenarios such as industrial PCB images, remote sensing and medical images. It is verified that CIS can achieve higher accuracy than the state-of-the-art method, while requiring less execution time. Moreover, by integrating SER into CIS, the proposed algorithm demonstrates excellent performance in further improving the anti-interference capability and positioning accuracy.

Trunk-branch Contrastive Network with Multi-view Deformable Aggregation for Multi-view Action Recognition

Multi-view action recognition aims to identify actions in a given multi-view scene. Traditional studies initially extracted refined features from each view, followed by implemented paired interaction and integration, but they potentially overlooked the critical local features in each view. When observing objects from multiple perspectives, individuals typically form a comprehensive impression and subsequently fill in specific details. Drawing inspiration from this cognitive process, we propose a novel trunk-branch contrastive network (TBCNet) for RGB-based multi-view action recognition. Distinctively, TBCNet first obtains fused features in the trunk block and then implicitly supplements vital details provided by the branch block via contrastive learning, generating a more informative and comprehensive action representation. Within this framework, we construct two core components: the multi-view deformable aggregation and the trunk-branch contrastive learning. MVDA employed in the trunk block effectively facilitates multi-view feature fusion and adaptive cross-view spatio-temporal correlation, where a global aggregation module is utilized to emphasize significant spatial information and a composite relative position bias is designed to capture the intra- and cross-view relative positions. Moreover, a trunk-branch contrastive loss is constructed between aggregated features and refined details from each view. By incorporating two distinct weights for positive and negative samples, a weighted trunk-branch contrastive loss is proposed to extract valuable information and emphasize subtle inter-class differences. The effectiveness of TBCNet is verified by extensive experiments on four datasets including NTU-RGB+D 60, NTU-RGB+D 120, PKU-MMD, and N-UCLA dataset. Compared to other RGB-based methods, our approach achieves state-of-the-art performance in cross-subject and cross-setting protocols.

MixSA: Training-free Reference-based Sketch Extraction via Mixture-of-Self-Attention

Current sketch extraction methods either require extensive training or fail to capture a wide range of artistic styles, limiting their practical applicability and versatility. We introduce Mixture-of-Self-Attention (MixSA), a training-free sketch extraction method that leverages strong diffusion priors for enhanced sketch perception. At its core, MixSA employs a mixture-of-self-attention technique, which manipulates self-attention layers by substituting the keys and values with those from reference sketches. This allows for the seamless integration of brushstroke elements into initial outline images, offering precise control over texture density and enabling interpolation between styles to create novel, unseen styles. By aligning brushstroke styles with the texture and contours of colored images, particularly in late decoder layers handling local textures, MixSA addresses the common issue of color averaging by adjusting initial outlines. Evaluated with various perceptual metrics, MixSA demonstrates superior performance in sketch quality, flexibility, and applicability. This approach not only overcomes the limitations of existing methods but also empowers users to generate diverse, high-fidelity sketches that more accurately reflect a wide range of artistic expressions.