Subjects: Computer Science >> Computer Software submitted time 2024-04-23
Abstract: We propose a method to guide Large Language Models (LLMs) in generating structured content adhering to specific conventions without fine-tuning. By utilizing coroutine-based content generation constraints through a pre-agreed context-free grammar (CFG), LLMs are directed during decoding to produce formal language compliant outputs. This enhances stability and consistency in generating target data structures, types, or instructions, reducing application development complexities. Experimentally, error rates of GPT-2 and Gemma exceed 95% for DSLs longer than 36 and 282 tokens, respectively. We introduce YieldLang, a coroutine-based DSL generation framework, and evaluate it with LLMs on various tasks including JSON and Mermaid flowchart generation. Compared to benchmarks, our approach improves accuracy by 1.09 to 11.6 times, with LLMs requiring only about 16.5% of the samples to generate JSON effectively. This enhances usability of LLM-generated content for computer programs.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Computer Software Subjects: Linguistics and Applied Linguistics >> Linguistics and Applied Linguistics submitted time 2024-04-21
Abstract: This slide presentation describes the research on Guiding Large Language Models to Generate Computer-Parsable Content in terms of Background, Motivation, Method, Effect, Prospect and Acknowledgements. For the full paper, please refer to: https://arxiv.org/abs/2404.05499
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Computer Software Subjects: Linguistics and Applied Linguistics >> Linguistics and Applied Linguistics submitted time 2024-04-07
Abstract: Large language models (LLMs) have demonstrated remarkable capabilities in learning patterns from massive text corpora, including word relationships, sentence structures, and even complex semantic and pragmatic information. However, it remains challenging to induce pre-trained language models to generate structured content that strictly follows specific conventions.We propose a scheme for guiding LLMs to generate highly usable content for computers without the need for fine-tuning and additional neural network inference, by introducing coroutine-based content generation constraints through a pre-agreed context-free grammar (CFG), which guides the autoregressive model Transformer to sample the correct tokens during its decoding phase to form a program-compliant form in the decoding phase of the autoregressive model Transformer to form a formal language that conforms to the program conventions. This will effectively improve the stability and consistency of LLMs in generating target data structures, types or instructions, and reduce the difficulty of application development and integration.We first verified that the error rate of models such as GPT-2 and Gemma reaches 95% when the length of the generated DSLs are greater than 36 and 282, respectively, through the experiment of matching bracket pairs , which illustrates the performance problem of some current LLMs in the generation of specific DSLs. We also present YieldLang, a coroutine-based DSL generation framework, and conduct experiments using LLMs on multiple task datasets, including tasks such as JSON, Mermaid flowchart, and function call expression generation. These experiments show that the approach in this paper improves its accuracy by a factor of 1.09 to 11.6 compared to the benchmarks, and in the best case is able to reduce the number of samples used by the LLMs to generate JSON to about 16.5% of the benchmarks, which will effectively improve the usability of the content generated by the LLMs for computer programs.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Computer Software submitted time 2024-01-07
Abstract: This thesis proposes a method for industrial process control loop fault diagnosis based on graph neural networks. By monitoring the output signals of loop sensors, the graph neural network can capture abnormal behaviors in the loop and automatically diagnose the type of loop faults. Experimental results demonstrate that the proposed method can efficiently detect loop faults and achieve high accuracy in both single and multiple fault scenarios. This method provides a reliable fault diagnosis solution for industrial process control, which has important practical significance and application value in actual industrial applications.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Computer Software submitted time 2022-12-07
Abstract:
For real parameter single objective optimization, Differential Evolution (DE) and Covariance Matrix Adaptation Evolution Strategy (CMA-ES) both perform powerfully. Nevertheless, in the field of real parameter single objective optimization, it is impossible for a given algorithm to perform well in all fitness landscapes. Practice has proved that ensemble of different algorithms may lead to improvement in solution. In this paper, based on two famous population-based metaheuristics - LSHADE-EpSin and HS-ES, we propose ensemble with successively executed constituent algorithms - HS-ES-DE. In our algorithm, HS-ES is replaced by L-SHADE-EpSin after stagnation is detected. Beside our HS-ES-DE, 12 population-based metaheuristics are involved in our experiments in which three benchmark test suites are employed. Experimental results show that our algorithm is very competitive.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Computer Software submitted time 2022-07-07
Abstract:
Intuitively there is drastic distinction between the “pure” decentralized block-chain systems like Defis and those that only utilizes block-chain as an enhancing technology but remains centralized with real-world business model and conventional technologies like database, application server etc. Our study explores extensively this distinction from a methodological point of view, classifies them into blockchain-complete and blockchain-partial, analyzes key features of the two types, and reveal the root cause of this distinction. We analyze the function or, in more strong words, the “ultimate purpose” of blockchain in the blockchain-partial systems, and present a conceptual model we named proof-chain that quite satisfactorily represented the general paradigm of blockchain in blockchain-partial systems. A universal tension between strength of proof-chain and privacy is then revealed and the zero-knowledge based proof-chain takes shape. Several case studies demonstrate the explaining power of our proof-chain methodology. We then apply proof-chain methodology to the analysis of the ecosystem of a collaborating group of blockchain-partial systems, representing the paradigm of public and private data domain whose border the proof-chain crosses. Finally, some derived guidelines from this methodology speaks usefulness of our methodology. " "
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Computer Software submitted time 2022-03-10
Abstract:
Distributed database system are widely used because of the rapid development of the Internet. With the ever-increasing demand, the boost performance and minimize resource and data contention are taken into consideration. A great distributed physical design, which determines where to place data, and which data item to replicate and partition, would help. This paper classification the development of physical design based on Michael’s work and its references in research problems, research methods and measurement methods. Finally we put forward some suggestions for future research.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Computer Software submitted time 2022-02-24
Abstract:
This paper focus on the researches of Maximal Independent Set (MIS). Based on reading and analysis of several recent papers, we divide the MIS problems into several classifications. The first is the classification based on the research objects, including the solution and maintenance of MIS; the second is the classification based on research methods, including serial, parallel, deterministic and randomized algorithms; the third is experimental analysis, including worst time complexity and expected time complexity.
Peer Review Status:Awaiting Review
Subjects: Mechanical Engineering >> Other Disciplines of Mechanical Engineering Subjects: Computer Science >> Computer Software submitted time 2022-02-16
Abstract:
"
"
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Computer Software submitted time 2022-01-01
Abstract: Serious noise affects the rendering of global illumination using Monte Carlo (MC) path tracing when insufficient samples are used. The two common solutions to this problem are filtering noisy inputs to generate smooth but biased results and sampling the MC integrand with a carefully crafted probability distribution function (PDF) to produce unbiased results. Both solutions benefit from an efficient incident radiance field sampling and reconstruction algorithm. This paper summarizes the latest advancement in using deep reinforcement learning in the adaptive sampling and reconstruction of the incident radiance field. "
Peer Review Status:Awaiting Review
Subjects: Mathematics >> Applied Mathematics Subjects: Computer Science >> Computer Software Subjects: Information Science and Systems Science >> Other Disciplines of Information Science and Systems Science submitted time 2021-10-11
Abstract: "
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Computer Software submitted time 2021-06-29
Abstract: " Evolution is the driving force behind the evolution of biological intelligence. Learning is the driving force behind human civilization. The combination of evolution and learning can form an entire natural world. Now, reinforcement learning has shown significant effects in many places. However, Currently, researchers in the field of optimization algorithms mainly focus on evolution strategies. And there is very little research on learning. Inspired by these ideas, this paper proposes a new particle swarm optimization algorithm Reinforcement learning based Ensemble particle swarm optimizer (RLEPSO) that combines reinforcement learning. The algorithm uses reinforcement learning for pre-training in the design phase to automatically find a more effective combination of parameters for the algorithm to run better and Complete optimization tasks faster. Besides, this algorithm integrates two robust particle swarm variants. And it sets the weight parameters for different algorithms to better adapt to the solution requirements of a variety of different optimization problems, which significantly improves the robustness of the algorithm. RLEPSO makes a certain number of sub-swarms to increase the probability of finding the global optimum and increasing the diversity of particle swarms. This proposed RLEPSO is evaluated on an optimization test functions benchmark set (CEC2013) with 28 functions and compared with other eight particle swarm optimization variants, including three state-of-the-art optimization algorithms. The results show that RLEPSO has better performance and outperforms all compared algorithms.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Computer Software submitted time 2021-03-01
Abstract: " This paper proposes a self-supervised low light image enhancement method based on deep learning, which can improve the image contrast and reduce noise at the same time to avoid the blur caused by pre-/post-denoising. The method contains two deep sub-networks, an Image Contrast Enhancement Network (ICE-Net) and a Re-Enhancement and Denoising Network (RED-Net). The ICE-Net takes the low light image as input and produces a contrast enhanced image. The RED-Net takes the result of ICE-Net and the low light image as input, and can re-enhance the low light image and denoise at the same time. Both of the networks can be trained with low light images only, which is achieved by a Maximum Entropy based Retinex (ME-Retinex) model and an assumption that noises are independently distributed. In the ME-Retinex model, a new constraint on the reflectance image is introduced that the maximum channel of the reflectance image conforms to the maximum channel of the low light image and its entropy should be the largest, which converts the decomposition of reflectance and illumination in Retinex model to a non-ill-conditioned problem and allows the ICE-Net to be trained with a self-supervised way. The loss functions of RED-Net are carefully formulated to separate the noises and details during training, and they are based on the idea that, if noises are independently distributed, after the processing of smoothing filters (\eg mean filter), the gradient of the noise part should be smaller than the gradient of the detail part. It can be proved qualitatively and quantitatively through experiments that the proposed method is efficient.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Computer Software submitted time 2017-04-10
Abstract:Traditional manifold learning algorithms often bear an assumption that the local neighborhood of any point on embedded manifold is roughly equal to the tangent space at that point without considering the curvature. The curvature indifferent way of manifold processing often makes traditional dimension reduction poorly neighborhood preserving. To overcome this drawback we propose a new algorithm called RF-ML to perform an operation on the manifold with help of Ricci flow before reducing the dimension of manifold.
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Computer Software submitted time 2017-03-09
Abstract:网络入侵检测与防御系统在当前的IP 网络安全领域中扮演着重要的角色,互联网流量的激增和单核处理器在数据包处理上存在的瓶颈,使得传统的运行于单核上的单线程网络入侵检测与防御系统已经远远不能满足网络发展的需求。为了解决这个问题,本文以主流单线程网络入侵检测与防御系统软件Snort 为基础,设计了一个基于软件流水的并行入侵检测系统pSnort,将传统的Snort 划分为2 个阶段,通过将其中最耗时的处理阶段并行化,以达到提升性能的目的。同时,通过程序设计,pSnort 避免了由于并行化而带来的严重的同步/互斥问题。经过试验,pSnort在Intel Quad-core Xeon 通用平台上可以获得超过1Gbps 的包处理速度。相对于传统的Snort,pSnort 最高能获得147%的性能提升以及2.5 倍加速比。
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Computer Software submitted time 2016-06-08
Abstract:随着高性能计算系统规模的不断扩大,节点失效愈加频发。传统的容错技术大都基于检查点(checkpoint)方式。但是,检查点技术的开销随着系统规模的扩大而不断增加,在百亿亿次(Exaflops)规模下其容错效率难以满足系统需求。算法失效恢复技术相比检查点方式具有更高的效率。然而,该技术依然基于停等模式。对于大规模系统,停等模式在很大程度上会影响程序的并行效率。本文提出了一种非停等的算法级容错策略——热替换策略。在程序运行过程中若发生节点失效,不用停等恢复失效节点上的数据,而用冗余节点替换失效节点,使计算能继续进行。最终的正确结果可以通过一个线性变换求出。为了论证方案的有效性,我们结合MPICH 的容错特性实现了容错的High Performance Linpack (HPL),并评估了方案的性能。实验结果表明,即使在小规模下,我们的方案的性能也明显优于算法失效恢复技术。
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Computer Software submitted time 2016-06-08
Abstract:在众核处理器系统中,片上网络常被用来提供高带宽、低延迟、高可靠的片上网络通信。为了减少网络拥塞、提高网络性能,流量平衡路由算法获得研究人员的广泛关注。流量平衡算法通常利用完全自适应路由算法来提供路径分集,而当前的完全自适应路由算法或者需要较多的虚通道或者假设一个保守的流控策略。一方面虚通道是比较昂贵的资源,另一方面保守的流控策略则有可能造成网络性能的下降。因此研究人员提出利用应用程序的流量信息来提升路由性能。这些算法在不使用虚通道的基础上可以针对不同的流量特性进行重构,从而实现路由自适应度的按需分配。按照使用的流量信息类型,流量感知的可重构路由算法可以分为离线和在线算法。离线算法需要事先知道程序的流量特征,因此他们大多针对应用程序定制的多核片上系统。在线算法则是根据在线收集的流量信息进行重构,因此可以用于通用处理器系统。本文将讨论最近国际上提出的两种著名的离线算法,并重点介绍本文作者在2011 年国际计算机体系结构大会(ISCA 11)上发表的基于算盘转向模型的在线可重构路由算法。
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Computer Software submitted time 2016-06-08
Abstract:本文报道了我们在CPU/ATI GPU 混合体系结构上优化双精度矩阵乘法(DGEMM)的工作。在真实应用中, CPU 与图形处理器(GPU)之间的数据传输是影响性能的关键因素。由于软件流水可以降低数据传输开销,我们提出了三种软件流水算法,分别是双缓存(Double Buffering)、数据重用(Data Reuse)和数据存储优化(Data Placement)。在AMD 公司的图形处理器(GPU)ATI HD5970 上,优化后DGEMM性能达到758 GFLOP/s,对应效率为82%,是ACML-GPU v1.1 性能的两倍。在Intel Westmere EP 和ATIHD5970 组成的异构系统上,性能达到844 GFLOP/s,效率为80%。我们进一步考察了多个CPU 和多个GPU上DGEMM 的扩展性,详细分析了体系结构方面的影响因素。分析表明,PCIe 总线和内存总线的竞争是异构系统上程序性能降低的重要影响因素。
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Computer Software submitted time 2016-06-08
Abstract:随着芯片内部处理器核数的增多,多核处理器逐渐有向众核方向发展的趋势。而众核这一全新的体系结构给计算机模拟带来了挑战。串行模拟已经难以满足速度的需求,必须充分利用现有并行宿主机的多核资源,在保证不损失模拟精度的前提下提升模拟速度。本文以众核和众核集群两种体系结构为例,说明并行模拟技术在计算机并行体系结构模拟中的必要性和可行性,在众核模拟中,做到精度不变,模拟速度提升10 倍;在众核集群模拟中,所模拟的处理器小核总数达到千核规模,并实现了混合的编程运行环境,为该结构的可扩展性测试提供了基础。
Peer Review Status:Awaiting Review