Understanding neural networks is challenging in part because of the dense, continuous nature of their hidden states. We explore whether we can train neural networks to have hidden states that are sparse, discrete, and more interpretable by quantizing their continuous features into what we call codebook features. Codebook features are produced by finetuning neural networks with vector quantization bottlenecks at each layer, producing a network whose hidden features are the sum of a small number of discrete vector codes chosen from a larger codebook. Surprisingly, we find that neural networks can operate under this extreme bottleneck with only modest degradation in performance. This sparse, discrete bottleneck also provides an intuitive way of controlling neural network behavior: first, find codes that activate when the desired behavior is present, then activate those same codes during generation to elicit that behavior. We validate our approach by training codebook Transformers on several different datasets. First, we explore a finite state machine dataset with far more hidden states than neurons. In this setting, our approach overcomes the superposition problem by assigning states to distinct codes, and we find that we can make the neural network behave as if it is in a different state by activating the code for that state. Second, we train Transformer language models with up to 410M parameters on two natural language datasets. We identify codes in these models representing diverse, disentangled concepts (ranging from negative emotions to months of the year) and find that we can guide the model to generate different topics by activating the appropriate codes during inference. Overall, codebook features appear to be a promising unit of analysis and control for neural networks and interpretability. Our codebase and models are open-sourced at //github.com/taufeeque9/codebook-features.
相關內容
Networking：IFIP International Conferences on Networking。
Explanation：國際網絡(luo)會議。
Publisher：IFIP。
SIT：
As optimization challenges continue to evolve, so too must our tools and understanding. To effectively assess, validate, and compare optimization algorithms, it is crucial to use a benchmark test suite that encompasses a diverse range of problem instances with various characteristics. Traditional benchmark suites often consist of numerous fixed test functions, making it challenging to align these with specific research objectives, such as the systematic evaluation of algorithms under controllable conditions. This paper introduces the Generalized Numerical Benchmark Generator (GNBG) for single-objective, box-constrained, continuous numerical optimization. Unlike existing approaches that rely on multiple baseline functions and transformations, GNBG utilizes a single, parametric, and configurable baseline function. This design allows for control over various problem characteristics. Researchers using GNBG can generate instances that cover a broad array of morphological features, from unimodal to highly multimodal functions, various local optima patterns, and symmetric to highly asymmetric structures. The generated problems can also vary in separability, variable interaction structures, dimensionality, conditioning, and basin shapes. These customizable features enable the systematic evaluation and comparison of optimization algorithms, allowing researchers to probe their strengths and weaknesses under diverse and controllable conditions.
Sentence classification is one of the basic tasks of natural language processing. Convolution neural network (CNN) has the ability to extract n-grams features through convolutional filters and capture local correlations between consecutive words in parallel, so CNN is a popular neural network architecture to dealing with the task. But restricted by the width of convolutional filters, it is difficult for CNN to capture long term contextual dependencies. Attention is a mechanism that considers global information and pays more attention to keywords in sentences, thus attention mechanism is cooperated with CNN network to improve performance in sentence classification task. In our work, we don't focus on keyword in a sentence, but on which CNN's output feature map is more important. We propose a Squeeze-and-Excitation Convolutional neural Network (SECNN) for sentence classification. SECNN takes the feature maps from multiple CNN as different channels of sentence representation, and then, we can utilize channel attention mechanism, that is SE attention mechanism, to enable the model to learn the attention weights of different channel features. The results show that our model achieves advanced performance in the sentence classification task.
The resolution is an important performance metric of near-field communication networks. In particular, the resolution of near field beamforming measures how effectively users can be distinguished in the distance-angle domain, which is one of the most significant features of near-field communications. In a comparison, conventional far-field beamforming can distinguish users in the angle domain only, which means that near-field communication yields the full utilization of user spatial resources to improve spectrum efficiency. In the literature of near-field communications, there have been a few studies on whether the resolution of near-field beamforming is perfect. However, each of the existing results suffers its own limitations, e.g., each is accurate for special cases only, and cannot precisely and comprehensively characterize the resolution. In this letter, a general analytical framework is developed to evaluate the resolution of near-field beamforming. Based on this derived expression, the impacts of parameters on the resolution are investigated, which can shed light on the design of the near-field communications, including the designs of beamforming and multiple access tequniques.
Community detection is the problem of identifying natural divi sions in networks. Efficient parallel algorithms for identifying such divisions is critical in a number of applications, where the size of datasets have reached significant scales. This technical report presents an optimized parallel implementation of Louvain, a high quality community detection method, for shared memory multicore systems. On a server equipped with dual 16-core Intel Xeon Gold 6226R processors, our Louvain, which we term as GVE-Louvain, outperforms Vite, Grappolo, and NetworKit Louvain by 50x, 22x, and 20x respectively - achieving a processing rate of 560M edges/s on a 3.8B edge graph. In addition, GVE-Louvain improves perfor mance at an average rate of 1.6x for every doubling of threads
Deep neural networks (DNNs) are vulnerable to adversarial perturbation, where an imperceptible perturbation is added to the image that can fool the DNNs. Diffusion-based adversarial purification focuses on using the diffusion model to generate a clean image against such adversarial attacks. Unfortunately, the generative process of the diffusion model is also inevitably affected by adversarial perturbation since the diffusion model is also a deep network where its input has adversarial perturbation. In this work, we propose MimicDiffusion, a new diffusion-based adversarial purification technique, that directly approximates the generative process of the diffusion model with the clean image as input. Concretely, we analyze the differences between the guided terms using the clean image and the adversarial sample. After that, we first implement MimicDiffusion based on Manhattan distance. Then, we propose two guidance to purify the adversarial perturbation and approximate the clean diffusion model. Extensive experiments on three image datasets including CIFAR-10, CIFAR-100, and ImageNet with three classifier backbones including WideResNet-70-16, WideResNet-28-10, and ResNet50 demonstrate that MimicDiffusion significantly performs better than the state-of-the-art baselines. On CIFAR-10, CIFAR-100, and ImageNet, it achieves 92.67\%, 61.35\%, and 61.53\% average robust accuracy, which are 18.49\%, 13.23\%, and 17.64\% higher, respectively. The code is available in the supplementary material.
The reliability of wireless Ad Hoc Networks (WANET) communication is much lower than wired networks. WANET will be impacted by node overload, routing protocol, weather, obstacle blockage, and many other factors, all those anomalies cannot be avoided. Accurate prediction of the network entirely stopping in advance is essential after people could do networking re-routing or changing to different bands. In the present study, there are two primary goals. Firstly, design anomaly events detection patterns based on Metamorphic Testing (MT) methodology. Secondly, compare the performance of evaluation metrics, such as Transfer Rate, Occupancy rate, and the Number of packets received. Compared to other studies, the most significant advantage of mathematical interpretability, as well as not requiring dependence on physical environmental information, only relies on the networking physical layer and Mac layer data. The analysis of the results demonstrates that the proposed MT detection method is helpful for automatically identifying incidents/accident events on WANET. The physical layer transfer Rate metric could get the best performance.
Since real-world objects and their interactions are often multi-modal and multi-typed, heterogeneous networks have been widely used as a more powerful, realistic, and generic superclass of traditional homogeneous networks (graphs). Meanwhile, representation learning (\aka~embedding) has recently been intensively studied and shown effective for various network mining and analytical tasks. In this work, we aim to provide a unified framework to deeply summarize and evaluate existing research on heterogeneous network embedding (HNE), which includes but goes beyond a normal survey. Since there has already been a broad body of HNE algorithms, as the first contribution of this work, we provide a generic paradigm for the systematic categorization and analysis over the merits of various existing HNE algorithms. Moreover, existing HNE algorithms, though mostly claimed generic, are often evaluated on different datasets. Understandable due to the application favor of HNE, such indirect comparisons largely hinder the proper attribution of improved task performance towards effective data preprocessing and novel technical design, especially considering the various ways possible to construct a heterogeneous network from real-world application data. Therefore, as the second contribution, we create four benchmark datasets with various properties regarding scale, structure, attribute/label availability, and \etc.~from different sources, towards handy and fair evaluations of HNE algorithms. As the third contribution, we carefully refactor and amend the implementations and create friendly interfaces for 13 popular HNE algorithms, and provide all-around comparisons among them over multiple tasks and experimental settings.
Ensembles over neural network weights trained from different random initialization, known as deep ensembles, achieve state-of-the-art accuracy and calibration. The recently introduced batch ensembles provide a drop-in replacement that is more parameter efficient. In this paper, we design ensembles not only over weights, but over hyperparameters to improve the state of the art in both settings. For best performance independent of budget, we propose hyper-deep ensembles, a simple procedure that involves a random search over different hyperparameters, themselves stratified across multiple random initializations. Its strong performance highlights the benefit of combining models with both weight and hyperparameter diversity. We further propose a parameter efficient version, hyper-batch ensembles, which builds on the layer structure of batch ensembles and self-tuning networks. The computational and memory costs of our method are notably lower than typical ensembles. On image classification tasks, with MLP, LeNet, and Wide ResNet 28-10 architectures, our methodology improves upon both deep and batch ensembles.
Deep neural networks (DNNs) are successful in many computer vision tasks. However, the most accurate DNNs require millions of parameters and operations, making them energy, computation and memory intensive. This impedes the deployment of large DNNs in low-power devices with limited compute resources. Recent research improves DNN models by reducing the memory requirement, energy consumption, and number of operations without significantly decreasing the accuracy. This paper surveys the progress of low-power deep learning and computer vision, specifically in regards to inference, and discusses the methods for compacting and accelerating DNN models. The techniques can be divided into four major categories: (1) parameter quantization and pruning, (2) compressed convolutional filters and matrix factorization, (3) network architecture search, and (4) knowledge distillation. We analyze the accuracy, advantages, disadvantages, and potential solutions to the problems with the techniques in each category. We also discuss new evaluation metrics as a guideline for future research.
Distant supervision can effectively label data for relation extraction, but suffers from the noise labeling problem. Recent works mainly perform soft bag-level noise reduction strategies to find the relatively better samples in a sentence bag, which is suboptimal compared with making a hard decision of false positive samples in sentence level. In this paper, we introduce an adversarial learning framework, which we named DSGAN, to learn a sentence-level true-positive generator. Inspired by Generative Adversarial Networks, we regard the positive samples generated by the generator as the negative samples to train the discriminator. The optimal generator is obtained until the discrimination ability of the discriminator has the greatest decline. We adopt the generator to filter distant supervision training dataset and redistribute the false positive instances into the negative set, in which way to provide a cleaned dataset for relation classification. The experimental results show that the proposed strategy significantly improves the performance of distant supervision relation extraction comparing to state-of-the-art systems.
小貼士
登錄享
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191