Wiki

In artificial intelligence, expert systems are computer systems that simulate the decision-making ability of human experts. Expert systems are the most important and active application field in artificial intelligence. They have achieved a major breakthrough in artificial intelligence from theoretical research to practical application, and from general reasoning strategy exploration to the use of specialized knowledge.

Zero copy technology means that when a computer performs an operation, the CPU does not need to copy data from one memory location to another specific area first.

The main purpose of Zero-Shot is to gain the ability to predict results without any training samples, the machine must recognize objects in classes that were not trained during training. Zero-shot learning is based on knowledge transfer, which is already included in the examples fed during training.

MMLU provides a way to test and compare various language models, such as OpenAI GPT-4, Mistral 7b, Google Gemini, and Anthropic Claude 2.

AIOps refers to the use of big data, advanced analytics, and machine learning to enhance the operational and functional workflows of IT teams.

Apptainer is a container system for high performance computing (HPC), formerly known as Singularity. It is used to build and run Linux containers, packaging software, libraries, and runtime compilers in isolated environments.

The Beowulf cluster is a high-performance parallel computer cluster structure, which is assembled using cheap personal computer hardware to achieve the best cost-effectiveness.

Human-Feedback Reinforcement Learning is an advanced method for training AI systems that combines reinforcement learning with human feedback.

Batch processing is to combine a series of commands or programs together in sequence and execute them in batches in a batch file.

Computational fluid dynamics applications run most efficiently on high performance computing (HPC) systems with high-throughput local storage, low-latency networks, and optimized CPUs.

Data Science (DS) aims to extract valuable information, insights and knowledge from large-scale data.

Field Programmable Gate Array A Field Programmable Gate Array (FPGA) is a semiconductor device based on a matrix of configurable logic blocks (CLBs) connected by programmable interconnects. It can be reprogrammed after fabrication based on the desired application or functional requirements.

In computer computing, Remote Direct Memory Access (RDMA) is a direct memory access technology that transfers data directly from the memory of one computer to another without the intervention of the operating systems of both computers.

Hardware acceleration refers to the process of assigning very computationally intensive tasks to specialized hardware for processing in a computer, which can reduce the workload of the central processing unit and is more efficient than software running solely on a general-purpose CPU.

Parallel computing is a sub-field of high performance computing (HPC). Relative to serial computing, it is a computing mode that improves computing efficiency by executing multiple tasks on multiple processors or computers simultaneously.

High-throughput computing (HTC) is defined as a type of computing that aims to use resources to run a large number of computing tasks in parallel.

The term High Performance Computing (HPC) emerged after the term "supercomputing" and is a field of computing that uses powerful computing resources to solve complex problems.

The Big Language Model is an artificial intelligence algorithm that uses a neural network with a large number of parameters to process and understand human language or text using self-supervised learning techniques.

Output modulation is a method of transforming output representation and causing disturbances, which is often used to increase the diversity of learners. It is to transform classification output into regression output and then construct individual learners.

Random forest is a versatile algorithm that contains multiple decision trees.

Random walk is a statistical model consisting of a series of random action trajectories, which is used to represent irregular changes.

Neural Machine Translation (NMT) is a machine translation framework based on pure neural networks. It uses neural networks to achieve end-to-end translation from source language to target language.

The Neural Turing Machine is a Turing Machine based on a neural network. It is inspired by the Turing Machine and can implement a machine algorithm for differential functions. It includes a neural network controller and external memory.

The same strategy means that the strategy for generating samples is the same as the strategy used when the network updates parameters. A typical example of the same strategy method is the SARAS algorithm.