论文检索

Yanbing Zhou, Xiangmou Qu, Chenlong You, Jiyang Zhou, Jingyue Tang, Xin Zheng, Chunmao Cai, Yingbo Wu

Prototype-based federated learning has emerged as a promising approach that shares lightweight prototypes to transfer knowledge among clients with data heterogeneity in a model-agnostic manner. However, existing methods often collect prototypes directly from local models, which inevitably introduce inconsistencies into representation learning due to the biased data distributions and differing model architectures among clients. In this paper, we identify that both statistical and model heterogeneity create a vicious cycle of representation inconsistency, classifier divergence, and skewed prototype alignment, which negatively impacts the performance of clients. To break the vicious cycle, we propose a novel framework named Federated Learning via Semantic Anchors (FedSA) to decouple the generation of prototypes from local representation learning. We introduce a novel perspective that uses simple yet effective semantic anchors serving as prototypes to guide local models in learning consistent representations. By incorporating semantic anchors, we further propose anchor-based regularization with margin-enhanced contrastive learning and anchor-based classifier calibration to correct feature extractors and calibrate classifiers across clients, achieving intra-class compactness and inter-class separability of prototypes while ensuring consistent decision boundaries. We then update the semantic anchors with these consistent and discriminative prototypes, which iteratively encourage clients to collaboratively learn a unified data representation with robust generalization. Extensive experiments under both statistical and model heterogeneity settings show that FedSA significantly outperforms existing prototype-based FL methods on various classification tasks.

Xun Zhou, Xingyu Wu, Liang Feng, Zhichao Lu, Kay Chen Tan

Transferable neural architecture search (TNAS) has been introduced to design efficient neural architectures for multiple tasks, to enhance the practical applicability of NAS in real-world scenarios. In TNAS, architectural knowledge accumulated in previous search processes is reused to warm up the architecture search for new tasks. However, existing TNAS methods still search in an extensive search space, necessitating the evaluation of numerous architectures. To overcome this challenge, this work proposes a novel transfer paradigm, i.e., design principle transfer. In this work, the linguistic description of various structural components' effects on architectural performance is termed design principles. They are learned from established architectures and then can be reused to reduce the search space tasks by discarding unpromising architectures. Searching in the refined search space can boost both the search performance and efficiency for new NAS tasks. To this end, a large language model (LLM)-assisted design principle transfer (LAPT) framework is devised. In LAPT, LLM is applied to automatically reason the design principles from a set of given architectures, and then a principle adaptation method is applied to refine these principles progressively based on the search results. Experimental results demonstrate that LAPT can beat the state-of-the-art TNAS methods on most tasks and achieve comparable performance on the remainder.

Xiaoling Zhou, Ou Wu, Nan Yang

Compared to conventional long-tail learning, which focuses on addressing class-wise imbalances, generalized long-tail (GLT) learning considers that samples within each class still conform to long-tailed distributions due to varying attributes, known as attribute imbalance. In the presence of such imbalance, the assumption of equivalence between the class-conditional probability densities of the training and testing sets is no longer tenable. Existing GLT approaches typically employ regularization techniques to avoid directly modeling the class-conditional probability density (CCPD) ratio between training and test data, leading to suboptimal performance. This study aims to directly estimate this ratio, for which a novel class-attribute aware logit-adjusted (CALA) loss incorporating both the CCPD ratio and the class priors is presented. Two new GLT learning methods, named Heuristic-CALA and Meta-CALA, are then proposed, which estimate the CCPD ratio in the CALA loss by leveraging the neighborhood information of samples. Extensive experiments across diverse scenarios susceptible to class and attribute imbalances showcase the state-of-the-art performance of Meta-CALA. Furthermore, while Heuristic-CALA exhibits inferior performance compared to Meta-CALA, it incurs only negligible additional training time compared to the Cross-Entropy loss, yet surpasses existing methods by a significant margin.

Qirui Zhou, Yuanbo Wen, Ruizhi Chen, Ke Gao, Weiqiang Xiong, Ling Li, Qi Guo, Yanjun Wu, Yunji Chen

As a crucial operator in numerous scientific and engineering computing applications, the automatic optimization of General Matrix Multiplication (GEMM) with full utilization of ever-evolving hardware architectures (e.g. GPUs and RISC-V) is of paramount importance. While Large Language Models (LLMs) can generate functionally correct code for simple tasks, they have yet to produce high-performance code. The key challenge resides in deeply understanding diverse hardware architectures and crafting prompts that effectively unleash the potential of LLMs to generate high-performance code. In this paper, we propose a novel prompt mechanism called QiMeng-GEMM which enables LLMs to comprehend the architectural characteristics of different hardware platforms and automatically search for the optimization combinations for GEMM. The key of QiMeng-GEMM is a set of informative, adaptive, and iterative meta-prompts. Based on this, a searching strategy for optimal combinations of meta-prompts is used to iteratively generate high-performance code. Extensive experiments conducted on 4 leading LLMs, various paradigmatic hardware platforms, and representative matrix dimensions unequivocally demonstrate QiMeng-GEMM’s superior performance in auto-generating optimized GEMM code. Compared to vanilla prompts, our method achieves a performance enhancement of up to 113×. Even when compared to human experts, our method can reach 115% of cuBLAS on NVIDIA GPUs and 211% of OpenBLAS on RISC-V CPUs. Notably, while human experts often take months to optimize GEMM, our approach reduces the development cost by over 240×.

Qingyang Zhou, Kangjie Lu, Meng Xu

Causal inconsistency arises when the underlying causal graphs captured by generative models like Normalizing Flows are inconsistent with those specified in causal models like Struct Causal Models. This inconsistency can cause unwanted issues including unfairness. Prior works to achieve causal consistency inevitably compromise the expressiveness of their models by disallowing hidden layers. In this work, we introduce a new approach: Causally Consistent Normalizing Flow (CCNF). To the best of our knowledge, CCNF is the first causally consistent generative model that can approximate any distribution with multiple layers. CCNF relies on two novel constructs: a sequential representation of SCMs and partial causal transformations. These constructs allow CCNF to inherently maintain causal consistency without sacrificing expressiveness. CCNF can handle all forms of causal inference tasks, including interventions and counterfactuals. Through experiments, we show that CCNF outperforms current approaches in causal inference. We also empirically validate the practical utility of CCNF by applying it to real-world datasets and show how CCNF addresses challenges like unfairness effectively.

Pengyang Zhou, Chaochao Chen, Weiming Liu, Xinting Liao, Wenkai Shen, Jiahe Xu, Zhihui Fu, Jun Wang, Wu Wen, Xiaolin Zheng

Federated graph learning (FGL) has emerged as a promising approach to enable collaborative training of graph models while preserving data privacy. However, current FGL methods overlook the out-of-distribution (OOD) shifts that occur in real-world scenarios. The distribution shifts between training and testing datasets in each client impact the FGL performance. To address this issue, we propose federated graph OOD generalization framework FedGOG, which includes two modules, i.e., diffusion data exploration (DDE) and latent embedding decorrelation (LED). In DDE, all clients jointly train score models to accurately estimate the global graph data distribution and sufficiently explore sample space using score-based graph diffusion with conditional generation. In LED, each client models a global invariant GNN and a personalized spurious GNN. LED aims to decorrelate spuriousness from invariant relationships by minimizing the mutual information between two categories of latent embeddings from different GNN models. Extensive experiments on six benchmark datasets demonstrate the superiority of FedGOG.

Lijing Zhou, Bingsheng Zhang, Ziyu Wang, Tianpei Lu, Qingrui Song, Su Zhang, Hongrui Cui, Yu Yu

Probabilistic truncation has been widely used in a broad range of privacy-preserving machine learning (PPML) platforms, such as EdaBits (Crypto 20), ABY 2.0 (Usenix 21), Crypten (NIPS 21), Piranha-Falcon (Usenix 22), and Bicoptor (S&P 23), etc. In this work, we examine the problems of common probabilistic truncation protocols in PPML, and propose solutions from the perspectives of accuracy and efficiency. With regard to accuracy, we found the recommended precision parameters in many existing works are incorrect, leading to extremely low inference accuracy. We conducted a thorough analysis of their open-source code and found that their errors were mainly caused by simplified implementation; more specifically, random numbers are not correctly sampled in probabilistic truncation protocols. Based on this, we provide a detailed theoretical analysis to validate our views. With regard to efficiency, we identify limitations in the state-of-the-art secure comparison, Bicoptor’s (S&P 2023) DReLU protocol, which relies on the probabilistic truncation and is heavily constrained by the security parameter to eliminate errors, significantly impacting its performance. To address these challenges, we introduce a non-interactive deterministic truncation technique, replacing the original probabilistic truncation. Additionally, we propose a new technique for speeding up the ReLU/DReLU evaluation, which can be applied to the other non-linear functions as well. When the input size of DReLU is reduced to 7 bits, we can speed up approximately 5x the ReLU protocols w.r.t. ABY3, ABY2.0, EdaBits, and Bicoptor without compromising model accuracy. The improved protocol can complete a ReLU evaluation within 2 rounds and 704 bits overall communication when the input/output is secretly shared over the 64-bit ring, which yields a 92% communication reduction on original Bicoptor. Compared to existing PPML platforms with GPU acceleration, our benchmark indicates a 10x improvement in the DReLU protocol, and a 6x improvement in the ReLU protocol over Piranha-Falcon and a 3.7x improvement over Bicoptor. As a result, the overall PPML model inference could be sped up by 3-4 times.

Junjie Zhou, Ke Zhu, Jianxin Wu

Knowledge Distillation (KD) is essential in transferring dark knowledge from a large teacher to a small student network, such that the student can be much more efficient than the teacher but with comparable accuracy. Existing KD methods, however, rely on a large teacher trained specifically for the target task, which is both very inflexible and inefficient. In this paper, we argue that a SSL-pretrained model can effectively act as the teacher and its dark knowledge can be captured by the coordinate system or linear subspace where the features lie in. We then need only one forward pass of the teacher, and then tailor the coordinate system (TCS) for the student network. Our TCS method is teacher-free and applies to diverse architectures, works well for KD and practical few-shot learning, allows cross-architecture distillation with large capacity gap. Experiments show that TCS achieves significantly higher accuracy than state-of-the-art KD methods, while only requiring roughly half of their training time and GPU memory costs.

Hanjing Zhou, Mingze Yin, Wei Wu, Mingyang Li, Kun Fu, Jintai Chen, Jian Wu, Zheng Wang

Multi-modality pre-training paradigm that aligns protein sequences and biological descriptions has learned general protein representations and achieved promising performance in various downstream applications. However, these works were still unable to replicate the extraordinary success of language-supervised visual foundation models due to the ineffective usage of aligned protein-text paired data and the lack of an effective function-informed pre-training paradigm. To address these issues, this paper curates a large-scale protein-text paired dataset called ProtAnno with a property-driven sampling strategy, and introduces a novel function-informed protein pre-training paradigm. Specifically, the sampling strategy determines selecting probability based on the sample confidence and property coverage, balancing the data quality and data quantity in face of large-scale noisy data. Furthermore, motivated by significance of the protein specific functional mechanism, the proposed paradigm explicitly model protein static and dynamic functional segments by two segment-wise pre-training objectives, injecting fine-grained information in a function-informed manner. Leveraging all these innovations, we develop ProtCLIP, a multi-modality foundation model that comprehensively represents function-aware protein embeddings. On 22 different protein benchmarks within 5 types, including protein functionality classification, mutation effect prediction, cross-modal transformation, semantic similarity inference and protein-protein interaction prediction, our ProtCLIP consistently achieves SOTA performance, with remarkable improvements of 75% on average in five cross-modal transformation benchmarks, 59.9% in GO-CC and 39.7% in GO-BP protein function prediction. The experimental results verify the extraordinary potential of ProtCLIP serving as the protein multi-modality foundation model.

Guanhao Zhou, Yuefeng Han, Xiufan Yu

This paper studies the prediction task of tensor-on-tensor regression in which both covariates and responses are multi-dimensional arrays (a.k.a., tensors) across time with arbitrary tensor order and data dimension. Existing methods either focused on linear models without accounting for possibly nonlinear relationships between covariates and responses, or directly employed black-box deep learning algorithms that failed to utilize the inherent tensor structure. In this work, we propose a Factor Augmented Tensor-on-Tensor Neural Network (FATTNN) that integrates tensor factor models into deep neural networks. We begin with summarizing and extracting useful predictive information (represented by the ``factor tensor'') from the complex structured tensor covariates, and then proceed with the prediction task using the estimated factor tensor as input of a temporal convolutional neural network. The proposed methods effectively handle nonlinearity between complex data structures, and improve over traditional statistical models and conventional deep learning approaches in both prediction accuracy and computational cost. By leveraging tensor factor models, our proposed methods exploit the underlying latent factor structure to enhance the prediction, and in the meantime, drastically reduce the data dimensionality that speeds up the computation. The empirical performances of our proposed methods are demonstrated via simulation studies and real-world applications to three public datasets. Numerical results show that our proposed algorithms achieve substantial increases in prediction accuracy and significant reductions in computational time compared to benchmark methods.

Guangchong Zhou, Zeren Zhang, Guoliang Fan

Exploration in cooperative multi-agent reinforcement learning (MARL) remains challenging for value-based agents due to the absence of an explicit policy. Existing approaches include individual exploration based on uncertainty towards the system and collective exploration through behavioral diversity among agents. However, the introduction of additional structures often leads to reduced training efficiency and infeasible integration of these methods. In this paper, we propose Adaptive exploration via Identity Recognition~(AIR), which consists of two adversarial components: a classifier that recognizes agent identities from their trajectories, and an action selector that adaptively adjusts the mode and degree of exploration. We theoretically prove that AIR can facilitate both individual and collective exploration during training, and experiments also demonstrate the efficiency and effectiveness of AIR across various tasks.

Changhai Zhou, Yuhua Zhou, Shiyang Zhang, Yibin Wang, Zekai Liu

Low-Rank Adaptation (LoRA) has become increasingly popular for efficiently fine-tuning large language models (LLMs) with minimal resources. However, traditional methods that serve multiple LoRA models independently result in redundant computation and low GPU utilization. This paper addresses these inefficiencies by introducing Dynamic Operator Optimization (Dop), an advanced automated optimization technique designed to dynamically optimize the Segmented Gather Matrix-Vector Multiplication (SGMV) operator based on specific scenarios. SGMV's unique design enables batching GPU operations for different LoRA models, significantly improving computational efficiency. The Dop approach leverages a Search Space Constructor to create a hierarchical search space, dividing the program space into high-level structural sketches and low-level implementation details, ensuring diversity and flexibility in operator implementation. Furthermore, an Optimization Engine refines these implementations using evolutionary search, guided by a cost model that estimates program performance. This iterative optimization process ensures that SGMV implementations can dynamically adapt to different scenarios to maintain high performance. We demonstrate that Dop can improve throughput by 1.30-1.46 times in a SOTA multi-tenant LoRA serving.

Ao Zhou, Bin Liu, Jin Wang, Grigorios Tsoumakas

The accuracy of deep neural networks is significantly influenced by the effectiveness of mini-batch construction during training. In single-label scenarios, such as binary and multi-class classification tasks, it has been demonstrated that batch selection algorithms preferring samples with higher uncertainty achieve better performance than difficulty-based methods. Although there are two batch selection methods tailored for multi-label data, none of them leverage important uncertainty information. Adapting the concept of uncertainty to multi-label data is not a trivial task, since there are two issues that should be tackled. First, traditional variance or entropy-based uncertainty measures ignore fluctuations of predictions within sliding windows and the importance of the current model state. Second, existing multi-label methods do not explicitly exploit the label correlations, particularly the uncertainty-based label correlations that evolve during the training process. In this paper, we propose an uncertainty-based multi-label batch selection algorithm. It assesses uncertainty for each label by considering differences between successive predictions and the confidence of current outputs, and further leverages dynamic uncertainty-based label correlations to emphasize instances whose uncertainty is synergistically expressed across multiple labels. Empirical studies demonstrate the effectiveness of our method in improving the performance and accelerating the convergence of various multi-label deep learning models.

Yan Zhong, Zhixin Yan, Yi Xie, Shibin Wu, Huaidong Zhang, Lin Shu, Peiru Zhou

Diabetic foot neuropathy (DFN) is a critical factor leading to diabetic foot ulcers, which is one of the most common and severe complications of diabetes mellitus (DM) and is associated with high risks of amputation and mortality. Despite its significance, existing datasets do not directly derive from plantar data and lack continuous, long-term foot-specific information. To advance DFN research, we have collected a novel dataset comprising continuous plantar pressure data to recognize diabetic foot neuropathy. This dataset includes data from 94 DM patients with DFN and 41 DM patients without DFN. Moreover, traditional methods divide datasets by individuals, potentially leading to significant domain discrepancies in some feature spaces due to the absence of mid-domain data. In this paper, we propose an effective domain adaptation method to address this proplem. We split the dataset based on convolutional feature statistics and select appropriate sub-source domains to enhance efficiency and avoid negative transfer. We then align the distributions of each source and target domain pair in specific feature spaces to minimize the domain gap. Comprehensive results validate the effectiveness of our method on both the newly proposed dataset for DFN recognition and an existing dataset.

Haoran Zheng, Jieming Shi, Renchi Yang

Graph similarity computation (GSC) is to calculate the similarity between one pair of graphs, which is a fundamental problem with fruitful applications in the graph community. In GSC, graph edit distance (GED) and maximum common subgraph (MCS) are the two most adopted similarity metrics, both of which are NP-hard to compute. Instead of calculating the exact values, state-of-the-art solutions resort to leveraging graph neural networks (GNNs) to learn data-driven models for the estimation of GED and MCS. Most of them are built on components involving node-level interactions crossing graphs, which engender vast computation overhead but are of little avail in effectiveness. Motivated by this, in the paper, we present GraSP, a simple yet effective GSC approach for GED and MCS prediction. More concretely, GraSP achieves high result efficacy through several key instruments: enhanced node features via positional encoding and a GNN model augmented by a gating mechanism, residual connections, as well as multi-scale pooling. Theoretically, GraSP can surpass the 1-WL test, indicating its high expressiveness. Empirically, extensive experiments comparing GraSP against 10 competitors on multiple widely adopted benchmark datasets showcase the superiority of GraSP over prior arts in terms of both effectiveness and efficiency.

Hao Zheng, Zhigang Hu, Liu Yang, Meiguang Zheng, Aikun Xu, Boyu Wang

Negative transfer (NF) is a critical challenge in personalized federated learning (pFL). Existing methods primarily focus on adapting local data distribution on the client side, which can only resist NF, rather than avoid NF itself. To tackle NF at its root, we investigate its mechanism through the lens of the global model, and argue that it is caused by update conflicts among clients during server aggregation. In light of this, we propose a conflict-free client update aggregation strategy (ConFREE), which enables us to avoid NF in pFL. Specifically, ConFREE guides the global update direction by constructing a conflict-free guidance vector through projection and utilizes the optimal local improvements of the worst-performing clients near the guidance vector to regularize server aggregation. This prevents the conflicting components of updates from transferring, achieving balanced updates across different clients. Notably, ConFREE is model-agnostic and can be straightforwardly adopted as a complement to enhance various existing NF-resistance methods implemented on the client side. Extensive experiments demonstrate substantial improvements to existing pFL algorithms by leveraging ConFREE.

Zhilin Zhao, Longbing Cao, Yuanyu Wan

We consider a general and realistic scenario involving non-stationary time series, consisting of several offline intervals with different distributions within a fixed offline time horizon, and an online interval that continuously receives new samples. For non-stationary time series, the data distribution in the current online interval may have appeared in previous offline intervals. We theoretically explore the feasibility of applying knowledge from offline intervals to the current online interval. To this end, we propose the Mixture of Online and Offline Experts (MOOE). MOOE learns static offline experts from offline intervals and maintains a dynamic online expert for the current online interval. It then adaptively combines the offline and online experts using a meta expert to make predictions for the samples received in the online interval. Specifically, we focus on theoretical analysis, deriving parameter convergence, regret bounds, and generalization error bounds to prove the effectiveness of the algorithm.

Zhe Zhao, Pengkun Wang, Haibin Wen, Shuang Wang, Liheng Yu, Yang Wang

Time series forecasting plays a crucial role in domains such as finance, healthcare, and climate science. However, as modern time series data become increasingly complex, featuring high dimensionality, intricate spatiotemporal dependencies, and multi-scale evolutionary patterns, traditional analytical methods and existing predictive models face significant challenges. Although Large Language Models (LLMs) excel in capturing long-range dependencies, they still struggle with multi-scale dynamics and seasonal patterns. Moreover, while LLMs' semantic representation capabilities are rich, they often lack explicit alignment with the numerical patterns and temporal structures of time series data, leading to limitations in predictive accuracy and interpretability. To address these challenges, this paper proposes a novel framework, STEM-LTS (Semantic-TEmporal Modeling for Large-scale Time Series). STEM-LTS enhances the ability to capture complex spatiotemporal dependencies by integrating time series decomposition techniques with LLM-based modeling. The semantic-temporal alignment mechanism within the framework significantly improves LLMs' ability to interpret and forecast time series data. Additionally, we develop an adaptive multi-task learning strategy to optimize the model's performance across multiple dimensions. Through extensive experiments on various real-world datasets, we demonstrate that STEM-LTS achieves significant improvements in prediction accuracy, robustness to noise, and interpretability. Our work not only advances LLM-based time series analysis but also offers new perspectives on handling complex temporal data.

Yusheng Zhao, Junyu Luo, Xiao Luo, Jinsheng Huang, Jingyang Yuan, Zhiping Xiao, Ming Zhang

Test-time adaptation aims to adapt a well-trained model to potential distribution shifts at test time using only unlabeled test data, without access to the original training data. While previous efforts mainly focus on a single modality, test-time distribution shift in the multi-modal setting is more complex and calls for new solutions. This paper tackles the problem of multi-modal test-time adaptation by proposing a novel method named Attention Bootstrapping with Principal Entropy Minimization (ABPEM). We observe that test-time distribution shift causes misalignment across modalities, leading to a large gap between intra-modality discrepancies (measured by self-attention) and inter-modality discrepancies (measured by cross-attention). We name this the attention gap. This attention gap widens with more severe distribution shifts, hindering effective modality fusion. To mitigate this attention gap and encourage better modality fusion, we propose attention bootstrapping that promotes cross-attention with the guidance of self-attention. Moreover, to reduce the gradient noise in the commonly-used entropy minimization, we adopt principal entropy minimization, a refinement of entropy minimization that reduces gradient noise by focusing on the principal parts of entropy, excluding less reliable gradient information. Extensive experiments on the benchmarks validate the effectiveness of the proposed ABPEM in comparison with competing baselines.

Youpeng Zhao, Ming Lin, Huadong Tang, Qiang Wu, Jun Wang

Generative Large Language Models (LLMs) stand as a revolutionary advancement in the modern era of artificial intelligence (AI). However, scaling down LLMs for resource-constrained hardware, such as Internet-of-Things (IoT) devices requires non-trivial efforts and domain knowledge. In this paper, we propose a novel information-entropy framework for designing mobile-friendly generative language models. The whole design procedure involves solving a mathematical programming (MP) problem, which can be done on the CPU within minutes, making it nearly zero-cost. We evaluate our designed models, termed MeRino, across fourteen NLP downstream tasks, showing their competitive performance against the state-of-the-art autoregressive transformer models under the mobile setting. Notably, MeRino achieves similar or better performance on both language modeling and zero-shot learning tasks, compared to the 350M parameter OPT while being 4.9x faster on NVIDIA Jetson Nano with 5.5x reduction in model size.