arXiv subjects

Jing Gao

Publications and source records attributed to Jing Gao.

At least 73 records · Page 4Linked to original sources

arXiv2024

FIARSE: Model-Heterogeneous Federated Learning via Importance-Aware Submodel Extraction

In federated learning (FL), accommodating clients' varied computational capacities poses a challenge, often limiting the participation of those with constrained resources in global model training. To address this issue, the concept of model heterogeneity through submodel extraction has emerged, offering a tailored solution that aligns the model's complexity with each client's computational capacity. In this work, we propose Federated Importance-Aware Submodel Extraction (FIARSE), a novel approach that dynamically adjusts submodels based on the importance of model parameters, thereby overcoming the limitations of previous static and dynamic submodel extraction methods. Compared to existing works, the proposed method offers a theoretical foundation for the submodel extraction and eliminates the need for additional information beyond the model parameters themselves to determine parameter importance, significantly reducing the overhead on clients. Extensive experiments are conducted on various datasets to showcase the superior performance of the proposed FIARSE.

cs.DC↗

arXiv2024

Model LEGO: Creating Models Like Disassembling and Assembling Building Blocks

With the rapid development of deep learning, the increasing complexity and scale of parameters make training a new model increasingly resource-intensive. In this paper, we start from the classic convolutional neural network (CNN) and explore a paradigm that does not require training to obtain new models. Similar to the birth of CNN inspired by receptive fields in the biological visual system, we draw inspiration from the information subsystem pathways in the biological visual system and propose Model Disassembling and Assembling (MDA). During model disassembling, we introduce the concept of relative contribution and propose a component locating technique to extract task-aware components from trained CNN classifiers. For model assembling, we present the alignment padding strategy and parameter scaling strategy to construct a new model tailored for a specific task, utilizing the disassembled task-aware components. The entire process is akin to playing with LEGO bricks, enabling arbitrary assembly of new models, and providing a novel perspective for model creation and reuse. Extensive experiments showcase that task-aware components disassembled from CNN classifiers or new models assembled using these components closely match or even surpass the performance of the baseline, demonstrating its promising results for model reuse. Furthermore, MDA exhibits diverse potential applications, with comprehensive experiments exploring model decision route analysis, model compression, knowledge distillation, and more. The code is available at https://github.com/jiaconghu/Model-LEGO.

cs.CV↗

arXiv2024

BlendFilter: Advancing Retrieval-Augmented Large Language Models via Query Generation Blending and Knowledge Filtering

Retrieval-augmented Large Language Models (LLMs) offer substantial benefits in enhancing performance across knowledge-intensive scenarios. However, these methods often face challenges with complex inputs and encounter difficulties due to noisy knowledge retrieval, notably hindering model effectiveness. To address this issue, we introduce BlendFilter, a novel approach that elevates retrieval-augmented LLMs by integrating query generation blending with knowledge filtering. BlendFilter proposes the blending process through its query generation method, which integrates both external and internal knowledge augmentation with the original query, ensuring comprehensive information gathering. Additionally, our distinctive knowledge filtering module capitalizes on the intrinsic capabilities of the LLM, effectively eliminating extraneous data. We conduct extensive experiments on three open-domain question answering benchmarks, and the findings clearly indicate that our innovative BlendFilter surpasses state-of-the-art baselines significantly.

cs.CL↗

arXiv2024

RoseLoRA: Row and Column-wise Sparse Low-rank Adaptation of Pre-trained Language Model for Knowledge Editing and Fine-tuning

Pre-trained language models, trained on large-scale corpora, demonstrate strong generalizability across various NLP tasks. Fine-tuning these models for specific tasks typically involves updating all parameters, which is resource-intensive. Parameter-efficient fine-tuning (PEFT) methods, such as the popular LoRA family, introduce low-rank matrices to learn only a few parameters efficiently. However, during inference, the product of these matrices updates all pre-trained parameters, complicating tasks like knowledge editing that require selective updates. We propose a novel PEFT method, which conducts \textbf{r}ow and c\textbf{o}lumn-wise spar\textbf{se} \textbf{lo}w-\textbf{r}ank \textbf{a}daptation (RoseLoRA), to address this challenge. RoseLoRA identifies and updates only the most important parameters for a specific task, maintaining efficiency while preserving other model knowledge. By adding a sparsity constraint on the product of low-rank matrices and converting it to row and column-wise sparsity, we ensure efficient and precise model updates. Our theoretical analysis guarantees the lower bound of the sparsity with respective to the matrix product. Extensive experiments on five benchmarks across twenty datasets demonstrate that RoseLoRA outperforms baselines in both general fine-tuning and knowledge editing tasks.

cs.CL↗

arXiv2024

The maximum spectral radius of $θ_{1,3,3}$-free graphs with given size

A graph $G$ is said to be $F$-free if it does not contain $F$ as a subgraph. A theta graph, say $θ_{l_1,l_2,l_3}$, is the graph obtained by connecting two distinct vertices with three internally disjoint paths of length $l_1, l_2, l_3$, where $l_1\leq l_2\leq l_3$ and $l_2\geq2$. Recently, Li, Zhao and Zou [arXiv:2409.15918v1] characterized the $θ_{1,p,q}$-free graph of size $m$ having the largest spectral radius, where $q\geq p\geq3$ and $p+q\geq2k+1\geq7$, and proposed a problem on characterizing the graphs with the maximum spectral radius among $θ_{1,3,3}$-free graphs. In this paper, we consider this problem and determine the maximum spectral radius of $θ_{1,3,3}$-free graphs with size $m$ and characterize the extremal graph. Up to now, all the graphs in $\mathcal{G}(m,θ_{1,p,q})$ which have the largest spectral radius have been determined, where $q\geq p\geq 2$.

math.CO↗

arXiv2024

SaySelf: Teaching LLMs to Express Confidence with Self-Reflective Rationales

Large language models (LLMs) often generate inaccurate or fabricated information and generally fail to indicate their confidence, which limits their broader applications. Previous work elicits confidence from LLMs by direct or self-consistency prompting, or constructing specific datasets for supervised finetuning. The prompting-based approaches have inferior performance, and the training-based approaches are limited to binary or inaccurate group-level confidence estimates. In this work, we present the advanced SaySelf, a training framework that teaches LLMs to express more accurate fine-grained confidence estimates. In addition, beyond the confidence scores, SaySelf initiates the process of directing LLMs to produce self-reflective rationales that clearly identify gaps in their parametric knowledge and explain their uncertainty. This is achieved by using an LLM to automatically summarize the uncertainties in specific knowledge via natural language. The summarization is based on the analysis of the inconsistency in multiple sampled reasoning chains, and the resulting data is utilized for supervised fine-tuning. Moreover, we utilize reinforcement learning with a meticulously crafted reward function to calibrate the confidence estimates, motivating LLMs to deliver accurate, high-confidence predictions and to penalize overconfidence in erroneous outputs. Experimental results in both in-distribution and out-of-distribution datasets demonstrate the effectiveness of SaySelf in reducing the confidence calibration error and maintaining the task performance. We show that the generated self-reflective rationales are reasonable and can further contribute to the calibration. The code is made public at https://github.com/xu1868/SaySelf.

cs.CL↗

arXiv2024

SHIELD: Evaluation and Defense Strategies for Copyright Compliance in LLM Text Generation

Large Language Models (LLMs) have transformed machine learning but raised significant legal concerns due to their potential to produce text that infringes on copyrights, resulting in several high-profile lawsuits. The legal landscape is struggling to keep pace with these rapid advancements, with ongoing debates about whether generated text might plagiarize copyrighted materials. Current LLMs may infringe on copyrights or overly restrict non-copyrighted texts, leading to these challenges: (i) the need for a comprehensive evaluation benchmark to assess copyright compliance from multiple aspects; (ii) evaluating robustness against safeguard bypassing attacks; and (iii) developing effective defense targeted against the generation of copyrighted text. To tackle these challenges, we introduce a curated dataset to evaluate methods, test attack strategies, and propose lightweight, real-time defense to prevent the generation of copyrighted text, ensuring the safe and lawful use of LLMs. Our experiments demonstrate that current LLMs frequently output copyrighted text, and that jailbreaking attacks can significantly increase the volume of copyrighted output. Our proposed defense mechanism significantly reduces the volume of copyrighted text generated by LLMs by effectively refusing malicious requests. Code is publicly available at https://github.com/xz-liu/SHIELD

cs.CL↗

arXiv2024

Enhanced Self-Checkout System for Retail Based on Improved YOLOv10

With the rapid advancement of deep learning technologies, computer vision has shown immense potential in retail automation. This paper presents a novel self-checkout system for retail based on an improved YOLOv10 network, aimed at enhancing checkout efficiency and reducing labor costs. We propose targeted optimizations to the YOLOv10 model, by incorporating the detection head structure from YOLOv8, which significantly improves product recognition accuracy. Additionally, we develop a post-processing algorithm tailored for self-checkout scenarios, to further enhance the application of system. Experimental results demonstrate that our system outperforms existing methods in both product recognition accuracy and checkout speed. This research not only provides a new technical solution for retail automation but offers valuable insights into optimizing deep learning models for real-world applications.

cs.CV↗

arXiv2024

Spectral methods on a triangle and W-systems

We present an overarching framework for stable spectral methods on a triangle, defined by a multivariate W-system and based on orthogonal polynomials on the triangle. Motivated by the Koornwinder orthogonal polynomials on the triangle, we introduce a Koornwinder W-system. Once discretised by this W-system, the resulting spatial differentiation matrix is skew symmetric, affording important advantages insofar as stability and conservation of structure are concerned. We analyse the construction of the differentiation matrix and matrix vector multiplication, demonstrating optimal computational cost. Numerical convergence is illustrated through experiments with different parameter choices. As a result, our method exhibits key characteristics of a practical spectral method, inclusive of rapid convergence, fast computation and the preservation of structure of the underlying partial differential equation.

math.NA↗

arXiv2024

FedBiOT: LLM Local Fine-tuning in Federated Learning without Full Model

Large language models (LLMs) show amazing performance on many domain-specific tasks after fine-tuning with some appropriate data. However, many domain-specific data are privately distributed across multiple owners. Thus, this dilemma raises the interest in how to perform LLM fine-tuning in federated learning (FL). However, confronted with limited computation and communication capacities, FL clients struggle to fine-tune an LLM effectively. To this end, we introduce FedBiOT, a resource-efficient LLM fine-tuning approach to FL. Specifically, our method involves the server generating a compressed LLM and aligning its performance with the full model. Subsequently, the clients fine-tune a lightweight yet important part of the compressed model, referred to as an adapter. Notice that as the server has no access to the private data owned by the clients, the data used for alignment by the server has a different distribution from the one used for fine-tuning by clients. We formulate the problem into a bi-level optimization problem to minimize the negative effect of data discrepancy and derive the updating rules for the server and clients. We conduct extensive experiments on LLaMA-2, empirically showing that the adapter has exceptional performance when reintegrated into the global LLM. The results also indicate that the proposed FedBiOT significantly reduces resource consumption compared to existing benchmarks, all while achieving comparable performance levels.

cs.LG↗

arXiv2024

Towards Comprehensive Multimodal Perception: Introducing the Touch-Language-Vision Dataset

Tactility provides crucial support and enhancement for the perception and interaction capabilities of both humans and robots. Nevertheless, the multimodal research related to touch primarily focuses on visual and tactile modalities, with limited exploration in the domain of language. Beyond vocabulary, sentence-level descriptions contain richer semantics. Based on this, we construct a touch-language-vision dataset named TLV (Touch-Language-Vision) by human-machine cascade collaboration, featuring sentence-level descriptions for multimode alignment. The new dataset is used to fine-tune our proposed lightweight training framework, STLV-Align (Synergistic Touch-Language-Vision Alignment), achieving effective semantic alignment with minimal parameter adjustments (1%). Project Page: https://xiaoen0.github.io/touch.page/.

cs.CV↗

arXiv2024

Touch100k: A Large-Scale Touch-Language-Vision Dataset for Touch-Centric Multimodal Representation

Touch holds a pivotal position in enhancing the perceptual and interactive capabilities of both humans and robots. Despite its significance, current tactile research mainly focuses on visual and tactile modalities, overlooking the language domain. Inspired by this, we construct Touch100k, a paired touch-language-vision dataset at the scale of 100k, featuring tactile sensation descriptions in multiple granularities (i.e., sentence-level natural expressions with rich semantics, including contextual and dynamic relationships, and phrase-level descriptions capturing the key features of tactile sensations). Based on the dataset, we propose a pre-training method, Touch-Language-Vision Representation Learning through Curriculum Linking (TLV-Link, for short), inspired by the concept of curriculum learning. TLV-Link aims to learn a tactile representation for the GelSight sensor and capture the relationship between tactile, language, and visual modalities. We evaluate our representation's performance across two task categories (namely, material property identification and robot grasping prediction), focusing on tactile representation and zero-shot touch understanding. The experimental evaluation showcases the effectiveness of our representation. By enabling TLV-Link to achieve substantial improvements and establish a new state-of-the-art in touch-centric multimodal representation learning, Touch100k demonstrates its value as a valuable resource for research. Project page: https://cocacola-lab.github.io/Touch100k/.

cs.RO↗

arXiv2024

LIDAO: Towards Limited Interventions for Debiasing (Large) Language Models

Large language models (LLMs) have achieved impressive performance on various natural language generation tasks. Nonetheless, they suffer from generating negative and harmful contents that are biased against certain demographic groups (e.g., female), raising severe fairness concerns. As remedies, prior works intervened the generation by removing attitude or demographic information, inevitably degrading the generation quality and resulting in notable \textit{fairness-fluency} trade-offs. However, it is still under-explored to what extent the fluency \textit{has to} be affected in order to achieve a desired level of fairness. In this work, we conduct the first formal study from an information-theoretic perspective. We show that previous approaches are excessive for debiasing and propose LIDAO, a general framework to debias a (L)LM at a better fluency provably. We further robustify LIDAO in adversarial scenarios, where a carefully-crafted prompt may stimulate LLMs exhibiting instruction-following abilities to generate texts with fairness issue appears only when the prompt is also taken into account. Experiments on three LMs ranging from 0.7B to 7B parameters demonstrate the superiority of our method.

cs.LG↗

arXiv2024

Array SAR 3D Sparse Imaging Based on Regularization by Denoising Under Few Observed Data

Array synthetic aperture radar (SAR) three-dimensional (3D) imaging can obtain 3D information of the target region, which is widely used in environmental monitoring and scattering information measurement. In recent years, with the development of compressed sensing (CS) theory, sparse signal processing is used in array SAR 3D imaging. Compared with matched filter (MF), sparse SAR imaging can effectively improve image quality. However, sparse imaging based on handcrafted regularization functions suffers from target information loss in few observed SAR data. Therefore, in this article, a general 3D sparse imaging framework based on Regulation by Denoising (RED) and proximal gradient descent type method for array SAR is presented. Firstly, we construct explicit prior terms via state-of-the-art denoising operators instead of regularization functions, which can improve the accuracy of sparse reconstruction and preserve the structure information of the target. Then, different proximal gradient descent type methods are presented, including a generalized alternating projection (GAP) and an alternating direction method of multiplier (ADMM), which is suitable for high-dimensional data processing. Additionally, the proposed method has robust convergence, which can achieve sparse reconstruction of 3D SAR in few observed SAR data. Extensive simulations and real data experiments are conducted to analyze the performance of the proposed method. The experimental results show that the proposed method has superior sparse reconstruction performance.

eess.IV↗

arXiv2024

Transformer in Touch: A Survey

The Transformer model, initially achieving significant success in the field of natural language processing, has recently shown great potential in the application of tactile perception. This review aims to comprehensively outline the application and development of Transformers in tactile technology. We first introduce the two fundamental concepts behind the success of the Transformer: the self-attention mechanism and large-scale pre-training. Then, we delve into the application of Transformers in various tactile tasks, including but not limited to object recognition, cross-modal generation, and object manipulation, offering a concise summary of the core methodologies, performance benchmarks, and design highlights. Finally, we suggest potential areas for further research and future work, aiming to generate more interest within the community, tackle existing challenges, and encourage the use of Transformer models in the tactile field.

cs.LG↗

arXiv2024

Evaluating the Factuality of Large Language Models using Large-Scale Knowledge Graphs

The advent of Large Language Models (LLMs) has significantly transformed the AI landscape, enhancing machine learning and AI capabilities. Factuality issue is a critical concern for LLMs, as they may generate factually incorrect responses. In this paper, we propose GraphEval to evaluate an LLM's performance using a substantially large test dataset. Specifically, the test dataset is retrieved from a large knowledge graph with more than 10 million facts without expensive human efforts. Unlike conventional methods that evaluate LLMs based on generated responses, GraphEval streamlines the evaluation process by creating a judge model to estimate the correctness of the answers given by the LLM. Our experiments demonstrate that the judge model's factuality assessment aligns closely with the correctness of the LLM's generated outputs, while also substantially reducing evaluation costs. Besides, our findings offer valuable insights into LLM performance across different metrics and highlight the potential for future improvements in ensuring the factual integrity of LLM outputs. The code is publicly available at https://github.com/xz-liu/GraphEval.

cs.CL↗

arXiv2024

Bifurcation on Fully Nonlinear Elliptic Equations and Systems

In this paper, we study the following fully nonlinear elliptic equations \begin{equation*} \left\{\begin{array}{rl} \left(S_{k}(D^{2}u)\right)^{\frac1k}=λf(-u) & in\quadΩ\\ u=0 & on\quad \partialΩ\\ \end{array} \right. \end{equation*} and coupled systems \begin{equation*} \left\{\begin{array}{rl} (S_{k}(D^{2}u))^\frac1k=λg(-u,-v) & in\quadΩ\\ (S_{k}(D^{2}v))^\frac1k=λh(-u,-v) & in\quadΩ\\ u=v=0 & on\quad \partialΩ\\ \end{array} \right. \end{equation*} dominated by $k$-Hessian operators, where $Ω$ is a $(k$-$1)$-convex bounded domain in $\mathbb{R}^{N}$, $λ$ is a non-negative parameter, $f:\left[0,+\infty\right)\rightarrow\left[0,+\infty\right)$ is a continuous function with zeros only at $0$ and $g,h:\left[0,+\infty\right)\times \left[0,+\infty\right)\rightarrow \left[0,+\infty\right)$ are continuous functions with zeros only at $(\cdot,0)$ and $(0,\cdot)$. We determine the interval of $λ$ about the existence, non-existence, uniqueness and multiplicity of $k$-convex solutions to the above problems according to various cases of $f,g,h$, which is a complete supplement to the known results in previous literature. In particular, the above results are also new for Laplacian and Monge-Ampère operators. We mainly use bifurcation theory, a-priori estimates, various maximum principles and technical strategies in the proof.

math.AP↗

arXiv2024

Towards Poisoning Fair Representations

Fair machine learning seeks to mitigate model prediction bias against certain demographic subgroups such as elder and female. Recently, fair representation learning (FRL) trained by deep neural networks has demonstrated superior performance, whereby representations containing no demographic information are inferred from the data and then used as the input to classification or other downstream tasks. Despite the development of FRL methods, their vulnerability under data poisoning attack, a popular protocol to benchmark model robustness under adversarial scenarios, is under-explored. Data poisoning attacks have been developed for classical fair machine learning methods which incorporate fairness constraints into shallow-model classifiers. Nonetheless, these attacks fall short in FRL due to notably different fairness goals and model architectures. This work proposes the first data poisoning framework attacking FRL. We induce the model to output unfair representations that contain as much demographic information as possible by injecting carefully crafted poisoning samples into the training data. This attack entails a prohibitive bilevel optimization, wherefore an effective approximated solution is proposed. A theoretical analysis on the needed number of poisoning samples is derived and sheds light on defending against the attack. Experiments on benchmark fairness datasets and state-of-the-art fair representation learning models demonstrate the superiority of our attack.

cs.LG↗