All Articles
306 articles in this section
A/B Testing (ML)
A/B testing in machine learning is a controlled experiment method that compares two or more model variants in production to determine which delivers superior performance on real-world business metrics.
Activation Function
A mathematical function applied to a neuron's output in a neural network that introduces non-linearity, enabling models to learn complex patterns beyond simple linear relationships.
Active Learning
Active learning is a machine learning paradigm in which the algorithm selectively queries a human annotator for labels on the most informative data points, minimising labelling effort while maximising model performance.
Adversarial Machine Learning
Adversarial machine learning is the study of attacks that exploit weaknesses in machine learning models, such as crafted inputs that cause misclassification, and of the defences designed to make models more robust.
Agentic AI
Agentic AI refers to artificial intelligence systems designed to pursue goals autonomously over extended time horizons by perceiving their environment, reasoning about actions, executing multi-step plans, and learning from outcomes without requiring continuous human instruction.
Agentic RAG
Agentic RAG is an approach to retrieval-augmented generation in which autonomous AI agents dynamically decide when, what, and how to retrieve information, applying planning, reflection, and tool use rather than following a fixed retrieve-then-generate pipeline.
AI Agents
Autonomous AI systems that perceive their environment, reason over goals, select and execute actions using external tools, and operate across multi-step tasks with minimal human intervention.
AI Alignment
AI alignment is the field of research dedicated to ensuring that artificial intelligence systems pursue goals, values, and behaviours that are consistent with human intentions.
AI Automation
AI automation combines artificial intelligence with robotic process automation and workflow orchestration to eliminate repetitive tasks, reduce errors, and free human workers for higher-value work.
AI Benchmarking
The systematic evaluation of AI systems using standardised datasets, tasks, and metrics to measure capability, compare models, and track progress across research and deployment contexts.
AI Bias
Systematic and unfair discrimination introduced into artificial intelligence systems through biased training data, flawed model design, or problematic deployment decisions, leading to unequal outcomes across demographic groups or categories.
AI Data Centres in Malaysia
AI data centres in Malaysia are large-scale computing facilities optimised for artificial intelligence training and inference, concentrated in Johor and the Klang Valley, that have made the country one of Southeast Asia's fastest-growing markets for AI infrastructure.
AI Drug Discovery
AI drug discovery applies machine learning, deep learning, and generative modelling to accelerate the identification, design, and optimisation of therapeutic compounds across the pharmaceutical pipeline.
AI Ethics
AI ethics is the branch of applied ethics addressing the moral dimensions of designing, deploying, and governing artificial intelligence systems — covering fairness, accountability, transparency, privacy, and safety.
AI Guardrails
AI guardrails are runtime safety mechanisms that validate, filter, and enforce policies on large language model inputs and outputs in production systems, preventing harmful content, data leakage, prompt injection, and off-topic behaviour.
AI in Islamic Finance Malaysia
AI in Islamic finance Malaysia refers to the application of artificial intelligence technologies — including machine learning, natural language processing, and generative AI — to Shariah-compliant financial products, institutions, and regulatory processes in Malaysia.
AI in Malaysian Agriculture
Artificial intelligence is transforming Malaysia's agricultural sector through precision farming, drone monitoring, yield prediction, and supply chain optimisation, with applications spanning palm oil, paddy, rubber, and aquaculture industries.
AI in Malaysian Banking
An overview of artificial intelligence adoption in Malaysian banks, covering fraud detection, customer service automation, credit decisioning, and the regulatory framework set by Bank Negara Malaysia.
AI in Malaysian Education
The adoption of artificial intelligence in Malaysia's education system spans school-level personalised learning, university AI programmes, educator upskilling, and government-led digital transformation initiatives under the MyDigital Blueprint.
AI in Malaysian Healthcare
The deployment of artificial intelligence across Malaysia's public and private healthcare systems for diagnostics, hospital operations, drug development, and population health, governed by Ministry of Health policy and PDPA requirements.
AI in Malaysian Legal Industry
AI in Malaysia's legal industry encompasses the adoption of machine learning, natural language processing, and generative AI tools by law firms, the judiciary, and legal service providers to automate research, drafting, and compliance tasks.
AI in Malaysian Logistics
AI in Malaysian logistics encompasses the application of artificial intelligence to freight forwarding, port operations, last-mile delivery, supply chain optimisation, and customs management across Malaysia's transport network.
AI in Malaysian Manufacturing
Artificial intelligence adoption in Malaysian manufacturing covers predictive maintenance, computer vision quality control, demand forecasting, and supply-chain optimisation across the electronics, automotive, and food sectors.
AI in Malaysian Retail
AI in Malaysian retail encompasses the deployment of machine learning, computer vision, and natural language processing across Malaysia's retail sector, including e-commerce platforms, brick-and-mortar stores, and omnichannel retail operations.
AI in Malaysian Telecommunications
The application of artificial intelligence by Malaysian telecommunications operators to optimise network operations, improve customer experience, detect fraud, and enable enterprise 5G and AI services.
AI in the Malaysian Palm Oil Industry
The application of artificial intelligence, computer vision, robotics, and predictive analytics to oil palm cultivation, harvesting, milling, and supply chain traceability in Malaysia.
AI Literacy
AI literacy is the set of knowledge, skills, and attitudes that enable individuals to understand, evaluate, and use artificial intelligence tools effectively and responsibly in personal, professional, and civic contexts.
AI Memory
AI memory refers to the mechanisms that allow artificial intelligence agents to retain, retrieve, and use information across interactions, extending capability beyond a single context window.
AI Music Generation
AI music generation is the use of machine learning models to compose, arrange, or produce music from text prompts or other inputs, spanning full songs with vocals, instrumental tracks, and sound design.
AI PC
A personal computer equipped with a dedicated Neural Processing Unit (NPU) designed to accelerate on-device artificial intelligence workloads locally, without requiring cloud connectivity, for tasks such as image generation, speech recognition, and language model inference.
AI Planning
AI planning is the discipline of automatically generating a sequence of actions that an intelligent agent can execute to move from an initial state to a goal, increasingly used inside LLM-based agents to decompose and reason about complex tasks.
AI Red Teaming
A structured adversarial evaluation practice in which testers attempt to elicit harmful, unsafe, or policy-violating behaviour from AI systems in order to surface risks before deployment.
AI Regulation in Malaysia
An overview of the Malaysian regulatory landscape governing artificial intelligence — covering PDPA, sectoral guidelines, national AI policy, and Malaysia's approach to the global AI governance debate.
AI Safety
AI safety is a field of research and practice concerned with the development of artificial intelligence systems that behave reliably, avoid harmful outputs, and remain aligned with human values, especially as systems become more capable.
AI Singapore
AI Singapore (AISG) is a national programme launched in 2017 to build Singapore's artificial intelligence ecosystem through research, talent development, and industry adoption, hosted by the National University of Singapore.
AI Thailand
AI Thailand refers to Thailand's national strategy, government programmes, and regulatory framework for developing and deploying artificial intelligence, as articulated in the National AI Strategy and Action Plan covering 2022 to 2027.
AI Video Generation
AI video generation refers to the automated creation of video content from text prompts, images, or other inputs using generative neural networks, enabling synthetic video production without cameras or traditional animation.
AI Watermarking
AI watermarking refers to techniques for embedding detectable signals into AI-generated content to establish provenance, enable detection, and support content authenticity verification across images, audio, video, and text.
AI21 Labs
An Israeli artificial intelligence company founded in 2017 that develops large language models, including the Jurassic and Jamba families, and enterprise agentic platforms such as Maestro.
AlphaFold
AlphaFold is an AI system developed by Google DeepMind that predicts the three-dimensional structure of proteins from their amino acid sequences, achieving accuracy comparable to experimental methods.
Amazon Bedrock
Amazon Bedrock is a fully managed AWS service that provides enterprise-grade access to over 100 foundation models from leading AI providers through a unified API, enabling organisations to build, customise, and scale generative AI applications without managing infrastructure.
Amazon SageMaker
Amazon SageMaker is a fully managed cloud platform from AWS that provides an integrated environment for building, training, and deploying machine learning models at scale, incorporating tools for data preparation, model development, MLOps, and generative AI.
Anomaly Detection
A class of machine learning techniques that identifies rare events, observations, or patterns that differ significantly from the majority of data, used for fraud, intrusion, and fault detection.
Anthropic
Anthropic is an American AI safety company and large language model developer founded in 2021 by former OpenAI researchers, best known for developing the Claude family of AI assistants and the Constitutional AI alignment technique.
Arize AI
Arize AI is an American ML observability and LLM evaluation platform that helps teams monitor, debug, and improve artificial intelligence models in production, offering both open-source and enterprise-grade tooling.
Artificial Intelligence
Artificial intelligence (AI) is the simulation of human intelligence processes by computer systems, encompassing learning, reasoning, problem-solving, perception, and language understanding.
ASEAN AI Governance
ASEAN AI governance refers to the regional frameworks, guidelines, and policy initiatives adopted by Southeast Asian nations to promote responsible, trustworthy, and interoperable artificial intelligence development and deployment.
Attention Mechanism
A neural network technique that enables models to dynamically weight the relevance of different parts of an input sequence when producing each output element, forming the core of transformer architectures.
Autoencoder
An autoencoder is a type of artificial neural network trained to reconstruct its input through a compressed internal representation, used for dimensionality reduction, feature learning, and anomaly detection.
AutoGen
AutoGen is an open-source multi-agent conversation framework developed by Microsoft Research that enables developers to build LLM applications where multiple AI agents communicate with each other to accomplish complex tasks collaboratively.
AutoML
AutoML (Automated Machine Learning) is the process of automating the selection, composition, and tuning of machine learning algorithms and pipelines, enabling practitioners to build effective models with reduced manual effort.
Autonomous Agents
Autonomous AI agents are software systems that use large language models as a reasoning core, enabling them to plan multi-step tasks, use external tools, maintain memory, and take actions to achieve goals with minimal human intervention.
Azure AI
Azure AI is Microsoft's integrated portfolio of artificial intelligence services hosted on the Azure cloud platform, encompassing pre-built cognitive APIs, a managed machine learning platform, large language model access, and enterprise AI development tools.
Backpropagation
Backpropagation is the primary algorithm for training neural networks, computing gradients of a loss function with respect to each weight by applying the chain rule of calculus in reverse through the network layers.
Batch Normalisation
Batch normalisation is a deep learning technique that normalises the activations of each layer within a mini-batch to accelerate training and improve model stability.
Bayesian Inference
Bayesian inference is a statistical method that uses Bayes' theorem to update the probability of a hypothesis as new evidence becomes available, providing a principled framework for reasoning under uncertainty.
BERT
BERT (Bidirectional Encoder Representations from Transformers) is a pre-trained transformer-based language model developed by Google that reads text bidirectionally to understand word context in natural language tasks.
BM25
BM25 (Best Matching 25) is a probabilistic ranking function used in information retrieval that scores documents based on query term frequency, inverse document frequency, and document length normalisation.
BNM AI Guidelines
Bank Negara Malaysia's AI guidelines provide a regulatory framework governing the responsible adoption of artificial intelligence in Malaysia's financial sector, covering risk management, governance, and accountability requirements for licensed institutions.
Canary Deployment
Canary deployment is a progressive model release strategy in which a new version is exposed to a small subset of production traffic, allowing teams to validate performance and catch failures before a full rollout.
Causal AI
Causal AI is an approach to artificial intelligence that incorporates causal reasoning into machine learning models, enabling them to go beyond correlation-based prediction to answer questions about interventions and counterfactual outcomes.
Chain-of-Thought Prompting
A prompt engineering technique that improves large language model reasoning on complex tasks by instructing the model to generate explicit intermediate reasoning steps before arriving at a final answer.
Chatbot
A chatbot is a software application designed to simulate human conversation through text or voice, ranging from simple rule-based systems to sophisticated AI assistants powered by large language models.
ChatGLM
A family of open-source bilingual (Chinese-English) large language models developed by Zhipu AI and Tsinghua University, known for strong reasoning capabilities, large context windows, and enterprise-grade open-weight releases under MIT licensing.
Chroma
An open-source vector database designed for embedding-based applications, optimised for developer ergonomics and increasingly for large-scale serverless retrieval through a 2025 Rust-core rewrite.
Claude (Language Model)
A family of large language models developed by Anthropic, designed with a focus on safety, helpfulness, and Constitutional AI training methods for enterprise and consumer use.
CLIP
CLIP (Contrastive Language-Image Pre-training) is a multimodal neural network model developed by OpenAI that learns visual concepts from natural language descriptions by jointly training an image encoder and a text encoder on 400 million image-text pairs.
Code Generation
AI code generation is the use of large language models to automatically produce, complete, or transform source code from natural language descriptions, enabling assisted and autonomous software development.
Cohere
Cohere is a Canadian AI company specialising in enterprise large language models, offering Command, Embed, and Rerank model families alongside secure deployment infrastructure designed for regulated industries.
Comet ML
Comet ML is a cloud-based MLOps platform for tracking machine learning experiments, managing model versions, monitoring production models, and evaluating large language model applications.
Computer Vision
Computer vision is the field of artificial intelligence that enables machines to interpret and act upon visual information from the world — including images, video, and depth data.
Constitutional AI
Constitutional AI is an alignment method developed by Anthropic that trains language models to follow a set of written ethical principles by using the model itself to critique and revise its own outputs, reducing dependence on human feedback for harmlessness.
Context Window
The maximum number of tokens — including the prompt, prior conversation, retrieved documents, and the model's own output — that a large language model can process in a single forward pass.
Continual Learning
Continual learning is a machine learning paradigm in which models incrementally acquire knowledge from sequential tasks or data streams without forgetting previously learned information, addressing the stability-plasticity trade-off inherent in neural networks.
Contrastive Learning
Contrastive learning is a self-supervised machine learning paradigm that trains models to produce similar representations for related data pairs and dissimilar representations for unrelated pairs, enabling powerful feature learning without labelled data.
Convolutional Neural Network
A convolutional neural network (CNN) is a type of deep neural network that uses convolutional layers to automatically learn spatial hierarchies of features from grid-structured data, most commonly images.
Core ML
Core ML is Apple's on-device machine learning framework that enables iOS, macOS, watchOS, and tvOS applications to integrate pre-trained models for tasks including image classification, natural language processing, and sound analysis.
Cosine Similarity
Cosine similarity is a measure of similarity between two non-zero vectors equal to the cosine of the angle between them, widely used to compare embeddings in search and machine learning.
CrewAI
CrewAI is an open-source Python framework for orchestrating multiple AI agents in role-based collaborative workflows, enabling teams of specialised agents to tackle complex tasks through structured coordination and communication.
Cross-Entropy Loss
Cross-entropy loss is the standard objective function for training classification models, measuring the divergence between a predicted probability distribution and the true distribution of labels.
CUDA
NVIDIA's parallel computing platform and programming model that lets developers use GPUs for general-purpose computation, underpinning most modern deep learning frameworks.
Cyberjaya
A planned technology city in Selangor, Malaysia, forming the nucleus of the Multimedia Super Corridor and home to MDEC, AI research initiatives, and a growing cluster of AI-ready data centres.
CyberSecurity Malaysia
CyberSecurity Malaysia is the national cyber security specialist agency under the Ministry of Science, Technology and Innovation, responsible for incident response, digital forensics, certification, and capacity building, including emerging risks from artificial intelligence.
DALL-E
DALL-E is a series of text-to-image generative AI models developed by OpenAI that create photorealistic and artistic images from natural language prompts using diffusion and language-vision alignment techniques.
Data Augmentation
A set of techniques that expand a training dataset by creating modified copies of existing examples, helping deep learning models generalise better and reducing overfitting.
Data Labelling
Data labelling is the process of attaching meaningful tags, classes, or annotations to raw data so that supervised machine learning models can learn to predict those labels on unseen examples.
Data Pipeline
A data pipeline is an automated sequence of processes that ingests, transforms, and delivers data from source systems to destination systems for analysis, machine learning, or operational use.
DataOps
DataOps is an engineering methodology that applies agile, DevOps, and lean manufacturing principles to data pipelines, aiming for rapid, reliable, and repeatable delivery of analytics and machine learning data.
DataRobot
An American enterprise AI platform company that provides automated machine learning, model deployment, monitoring, and generative AI tools, headquartered in Boston, Massachusetts.
Deep Learning
Deep learning is a subfield of machine learning that uses multi-layered artificial neural networks to learn hierarchical representations from data, enabling state-of-the-art performance across vision, language, and speech tasks.
DeepSeek
A Chinese artificial intelligence company founded in 2023, known for developing open-source large language models including DeepSeek-R1 and DeepSeek-V3 that achieved performance competitive with leading Western AI systems.
DeepSpeed
DeepSpeed is an open-source deep learning optimisation library developed by Microsoft that enables efficient distributed training and inference of large-scale neural networks through memory and compute optimisations.
Differential Privacy
Differential privacy is a mathematical framework for analysing data that guarantees the output of a computation reveals little about any single individual, achieved by adding calibrated random noise to limit each record's influence.
Diffusion Model
A class of generative AI models that learn to reverse a gradual noise-addition process, enabling the generation of high-quality images, audio, and video from random noise guided by text or other conditioning signals.
Direct Preference Optimization
Direct Preference Optimization (DPO) is a stable, computationally efficient algorithm for aligning large language models with human preferences by directly optimising a policy from comparison data, without training a separate reward model or using reinforcement learning.
Domain Adaptation
Domain adaptation is a machine learning technique that transfers a model trained on a labelled source domain to perform effectively on a related but distinct target domain with limited or no labelled target data, addressing distribution shift between domains.
Doubao
A suite of large language models and consumer AI assistant developed by ByteDance, the parent company of TikTok, reaching 159 million monthly active users and embedded across ByteDance's content, social, and device ecosystems.
Dropout
A regularisation technique in deep learning that randomly deactivates neurons during training, preventing co-adaptation and improving generalisation. Introduced by Hinton and colleagues in 2012 and formalised in 2014.
Edge AI
Edge AI is the deployment of artificial intelligence algorithms and inference workloads directly on local devices or edge computing nodes rather than in centralised cloud data centres, enabling low-latency, privacy-preserving, and bandwidth-efficient AI applications.
ElevenLabs
ElevenLabs is an AI audio research and deployment company founded in 2022 that develops text-to-speech, voice cloning, dubbing, and conversational voice agent technologies based on proprietary deep learning models.
Embedding
An embedding is a dense numerical vector representation of data — such as text, images, or audio — that encodes semantic meaning in a continuous high-dimensional space, enabling machine learning models to measure similarity and relationships.
Encoder-Decoder Architecture
A neural network design pattern that compresses an input sequence into an internal representation using an encoder, and then generates an output sequence from that representation using a decoder, foundational to machine translation, summarisation, and many other sequence-to-sequence tasks.
ERNIE Bot
A large language model and conversational AI assistant developed by Baidu, built on the ERNIE (Enhanced Representation through Knowledge Integration) foundation model series and integrated across Baidu's search, cloud, and enterprise platforms.
EU AI Act
The EU AI Act is the world's first comprehensive legal framework regulating artificial intelligence, classifying AI systems by risk level and imposing obligations on developers and deployers operating in the European Union.
Explainable AI
Explainable AI (XAI) refers to methods and techniques that make the decisions and predictions of artificial intelligence systems interpretable and understandable to human users, addressing the opacity of complex machine learning models.
Face Recognition
Face recognition is a biometric technology that identifies or verifies individuals by analysing facial features from images or video, widely used in security, banking, and immigration.
FAISS
FAISS (Facebook AI Similarity Search) is an open-source library developed by Meta AI for efficient similarity search and clustering of dense vectors at scale, enabling fast nearest-neighbour retrieval across millions or billions of embeddings.
Falcon LLM
A family of open-weight large language models developed by the Technology Innovation Institute (TII) in Abu Dhabi, released under permissive licenses and used widely across enterprise and research applications.
Feature Store
A centralised data platform for storing, serving, and managing machine learning features so that they can be reused consistently across training and online inference.
Federated Learning
Federated learning is a machine learning paradigm in which a model is trained across multiple decentralised devices or servers holding local data, without exchanging the raw data itself, preserving privacy while enabling collaborative model improvement.
Few-Shot Learning
Few-shot learning is a machine learning paradigm in which a model learns to perform new tasks or recognise new classes from only a small number of labelled training examples, often just one to five samples per class.
Fine-Tuning
The process of further training a pre-trained machine learning model on a smaller, task-specific dataset to adapt its weights for a particular domain, task, or desired behaviour.
Flash Attention
FlashAttention is an IO-aware exact attention algorithm that restructures the standard attention computation into memory-efficient tiled blocks, dramatically reducing GPU memory usage and wall-clock time for transformer models on long sequences.
Foundation Model
A large-scale AI model pretrained on broad, diverse datasets and designed to be adapted to a wide range of downstream tasks through fine-tuning, prompting, or retrieval augmentation.
Fraud Detection
Fraud detection is the application of data analysis, machine learning, and AI to identify deceptive or unauthorised transactions, activities, and behaviours in financial, digital, and commercial systems.
Function Calling
Function calling is the structured mechanism by which a large language model returns a JSON-formatted invocation of a named function with typed arguments, enabling reliable integration of LLMs with external systems.
Gaussian Process
A non-parametric Bayesian model that defines a distribution over functions, widely used in regression, optimisation, and uncertainty quantification.
Gemini
Gemini is a family of multimodal large language models developed by Google DeepMind, designed to natively process and generate text, code, images, audio, and video across a range of model sizes.
Gemma
Gemma is a family of open-weight large language models developed by Google DeepMind, built on similar technology to the Gemini series and available for deployment on hardware ranging from laptops to cloud infrastructure.
Generative Adversarial Network
A generative adversarial network (GAN) is a class of machine learning framework in which two neural networks, a generator and a discriminator, compete against each other to produce synthetic data indistinguishable from real examples.
Generative AI
Generative AI refers to artificial intelligence systems capable of producing new content — text, images, audio, video, or code — by learning the underlying distribution of training data.
Google DeepMind
Google DeepMind is an AI research laboratory owned by Alphabet Inc., formed in 2023 by the merger of Google Brain and DeepMind, and responsible for developing foundational AI systems including the Gemini family of models, AlphaFold, and AlphaGo.
Google Vertex AI
Google Vertex AI is a unified machine learning platform on Google Cloud that consolidates data preparation, model training, deployment, and monitoring for both custom-built models and Google's foundation models including Gemini.
GPT-4
GPT-4 is a large multimodal language model developed by OpenAI, released in March 2023, that accepts both image and text inputs and demonstrates human-level performance on numerous professional and academic benchmarks.
GPU Cluster
A GPU cluster is a networked group of servers, each containing one or more graphics processing units, purpose-built to accelerate parallel computation workloads such as deep learning training and large-scale AI inference.
Gradient Boosting
A machine learning ensemble technique that builds predictive models sequentially, where each new model corrects the errors of its predecessors using gradient descent optimisation.
Gradient Descent
Gradient descent is an iterative optimisation algorithm that minimises a loss function by repeatedly updating model parameters in the direction of the steepest descent, as defined by the negative gradient.
Graph Neural Network
A class of deep learning models designed to operate on graph-structured data, enabling nodes to aggregate and propagate information across their neighbourhoods through a message-passing mechanism.
Green AI
Green AI is an approach to artificial intelligence that prioritises energy efficiency and environmental sustainability across the model lifecycle, aiming to reduce the carbon footprint, power, and water consumption of training and inference.
Grok
Grok is a series of large language models developed by xAI, Elon Musk's AI company, featuring real-time web integration, advanced reasoning modes, and deep tool-use capabilities.
Groq
Groq is an American AI inference company that developed the Language Processing Unit (LPU), a custom silicon architecture optimised for high-throughput, low-latency inference of large language models using on-chip SRAM rather than external DRAM.
Hallucination (AI)
A phenomenon in which an artificial intelligence system generates output that is factually incorrect, fabricated, or unsupported by its input, while presenting it with apparent confidence.
Helicone
Helicone is an open-source LLM observability and gateway platform that enables developers to monitor, debug, and optimise large language model applications in production with minimal integration effort.
Hidden Markov Model
A statistical model that represents systems with unobservable (hidden) states that emit observable outputs, used widely in speech recognition, bioinformatics, and time-series analysis.
HRD Corp AI Training in Malaysia
HRD Corp, the Human Resource Development Corporation of Malaysia, funds employer-sponsored training in artificial intelligence, data science, and digital skills through a statutory levy collected from registered Malaysian employers.
Hugging Face
An American AI company and open-source platform that hosts machine learning models, datasets, and applications, widely described as the "GitHub of machine learning" for its role as the central repository of the open AI community.
Hunyuan
A family of large language models developed by Tencent, integrated across WeChat, QQ, and Tencent Cloud, offering multimodal capabilities including text, image, video, voice, and 3D generation through a unified omni-modal architecture.
Hybrid Search
Hybrid search is a retrieval technique that combines sparse keyword-based search (typically BM25) with dense vector semantic search to achieve superior recall and precision over either method alone.
Hyperparameter Tuning
The process of selecting optimal configuration values for a machine learning model's external parameters using methods such as grid search, random search, and Bayesian optimisation.
IBM watsonx
IBM's enterprise AI and data platform, comprising watsonx.ai for model development, watsonx.data as an open lakehouse, and watsonx.governance for AI risk and compliance management.
ILMU (Malaysian Large Language Model)
ILMU is Malaysia's first homegrown multimodal large language model, developed by YTL AI Labs to understand and generate Bahasa Melayu, Manglish and regional dialects across text, voice and vision.
Image Segmentation
A computer vision task that partitions an image into meaningful regions by assigning a class label to every pixel, enabling pixel-level understanding of visual scenes.
In-Context Learning
In-context learning is the ability of large language models to perform new tasks by conditioning on examples or instructions provided within the input prompt, without updating model weights.
Inference (Machine Learning)
Inference is the phase in which a trained machine learning model is used to generate predictions or outputs from new input data, distinct from the earlier training phase.
Instruction Tuning
Instruction tuning is a supervised fine-tuning technique that trains large language models on datasets of instruction-response pairs, enabling models to follow natural language directions and generalise to unseen tasks in a zero-shot or few-shot setting.
JAX
JAX is an open-source numerical computing library from Google that combines NumPy-style array programming with automatic differentiation and just-in-time compilation, used to train large-scale machine learning models on GPUs and TPUs.
K-Means Clustering
K-means clustering is an unsupervised machine learning algorithm that partitions a dataset into k groups by minimising the sum of squared distances between data points and their assigned cluster centroids.
Kimi
A conversational AI assistant and long-context large language model developed by Moonshot AI, a Beijing startup, known for its industry-leading context window lengths and strong performance on agentic reasoning tasks.
Kling AI
A family of generative AI video models developed by Kuaishou Technology in China, capable of producing photorealistic short-form video with synchronised audio from text or image prompts.
Knowledge Distillation
Knowledge distillation is a model compression technique in which a smaller student neural network is trained to replicate the behaviour of a larger, more capable teacher model, enabling deployment of efficient models that approximate teacher-level performance.
Knowledge Graph
A structured knowledge representation that encodes entities and their relationships as a directed labelled graph, enabling machines to reason over interconnected facts across diverse domains.
KV Cache
A KV cache (key-value cache) is a memory optimisation used in transformer inference that stores pre-computed key and value tensors from the attention mechanism, eliminating redundant recomputation when generating tokens sequentially.
Labelbox
Labelbox is an American AI data labeling and model evaluation platform that enables organisations to annotate training datasets, manage labeling workflows, and curate high-quality data for machine learning development.
LangChain
LangChain is an open-source framework for building applications powered by large language models, providing composable abstractions for chaining LLM calls with tools, memory, and data retrieval in Python and JavaScript.
Langfuse
Langfuse is an open-source LLM engineering platform that provides observability, tracing, prompt management, evaluation, and dataset tooling for teams building applications on top of large language models.
LangGraph
LangGraph is an open-source framework for building stateful, multi-actor AI agent applications using graph-based workflows, extending the LangChain ecosystem with cycle-capable execution and persistent state management.
LangSmith
LangSmith is an observability, tracing, and evaluation platform from LangChain for debugging, monitoring, and continuously improving large language model and AI agent applications in production.
Large Language Models
Large language models (LLMs) are AI systems trained on vast corpora of text to predict and generate natural language. They underpin modern chatbots, code assistants, and generative AI applications.
Layer Normalisation
Layer normalisation is a technique that normalises the inputs across the features of a single training example, stabilising and accelerating the training of deep neural networks, especially transformers.
Llama
Llama is a family of open-weight large language models developed by Meta AI, released under a permissive licence that allows researchers and developers to freely download, fine-tune, and deploy the models for both research and commercial use.
LlamaIndex
LlamaIndex is an open-source Python and TypeScript framework for building retrieval-augmented and agentic AI applications over private data sources.
Long Short-Term Memory (LSTM)
Long Short-Term Memory is a recurrent neural network architecture designed to learn long-range dependencies in sequential data by using gating mechanisms to control information flow.
LoRA (Low-Rank Adaptation)
LoRA is a parameter-efficient fine-tuning technique that adapts large pre-trained models by injecting small trainable low-rank matrices into transformer layers, drastically reducing the number of trainable parameters without sacrificing performance.
Machine Learning
Machine learning is a subfield of artificial intelligence in which systems improve their performance on tasks through experience — by automatically learning patterns from data rather than following explicitly programmed rules.
Machine Translation
Machine translation is the automated conversion of text or speech from one natural language into another using rule-based, statistical, or neural systems.
Malaysia AI Governance Framework
The Malaysia AI Governance Framework is a national set of guidelines, standards, and forthcoming legislation that defines responsible AI development and deployment in Malaysia.
Malaysia AI Talent
Malaysia AI talent refers to the workforce of AI engineers, data scientists, ML practitioners, and researchers in Malaysia, along with the government and industry initiatives aimed at developing and attracting AI expertise to meet the demands of the digital economy.
MaLLaM (Malaysia Large Language Model)
MaLLaM is a family of large language models developed by Malaysian startup Mesolitica, pretrained from scratch on Malay-language data to understand Malaysian dialects, colloquialisms, and regional languages.
Mamba (Structured State Space Model)
Mamba is a selective state space model architecture that achieves linear-time sequence modelling, offering a computationally efficient alternative to the Transformer for long-context tasks.
Markov Decision Process
A Markov decision process is a mathematical framework for modelling sequential decision-making in which outcomes are partly random and partly under the control of a decision-maker.
MDEC — Malaysia Digital Economy Corporation
MDEC (Malaysia Digital Economy Corporation) is the Malaysian government agency tasked with driving the country's digital economy — including AI adoption, digital talent development, and tech-sector FDI.
Meta AI
Meta AI is the artificial intelligence research division and product brand of Meta Platforms, responsible for the Llama family of open-weight language models and integrated AI assistants across Facebook, Instagram, WhatsApp, and Messenger.
Meta-Learning
A machine learning paradigm in which models learn how to learn, acquiring inductive biases across a distribution of tasks so they can adapt rapidly to new tasks with minimal data.
Microsoft Copilot
Microsoft Copilot is an AI-powered assistant integrated across Microsoft's product ecosystem — including Windows, Microsoft 365, Edge, and Azure — using large language models to assist with writing, coding, data analysis, and task automation.
Midjourney
An independent AI research lab and image generation service that produces images and video from natural-language text prompts, accessible primarily through Discord and a web application.
Milvus
Milvus is an open-source, cloud-native vector database built for high-performance approximate nearest neighbour search over massive embedding datasets, widely used in retrieval-augmented generation and semantic search.
MIMOS Berhad
MIMOS Berhad is Malaysia's national applied research and development centre under MOSTI, advancing technology platforms in artificial intelligence, semiconductors, and quantum computing.
MiniMax
A Chinese AI company and model developer known for the MiniMax-M1 and M2 large language models featuring ultra-long context windows of up to 4 million tokens, strong agentic performance, and open MIT-licensed releases.
Mistral AI
Mistral AI is a French artificial intelligence company founded in 2023 that develops and releases open-weight and proprietary large language models, notable for its competitive performance-to-efficiency ratio and commitment to open-source distribution.
Mixtral
Mixtral is a family of open-weight sparse mixture-of-experts large language models developed by Mistral AI, comprising Mixtral 8x7B and Mixtral 8x22B, released under the Apache 2.0 licence.
Mixture of Experts
Mixture of Experts (MoE) is a machine learning architecture in which a model routes each input to a small subset of specialised sub-networks called experts, enabling large model capacity at a fraction of the compute cost.
MLflow
An open-source platform for managing the end-to-end machine learning lifecycle, including experiment tracking, model packaging, a model registry, and deployment.
MLOps
A set of practices and tools that combine machine learning, DevOps, and data engineering to automate and operationalise the full lifecycle of ML models from development through production deployment and monitoring.
Model Cards
Model cards are structured documentation sheets accompanying machine learning models that disclose intended uses, performance characteristics, training data, limitations, and ethical considerations.
Model Compression
Model compression is a set of techniques that reduce the size, memory footprint, and computational cost of machine learning models while preserving predictive accuracy, enabling deployment on resource-constrained hardware.
Model Context Protocol
The Model Context Protocol (MCP) is an open standard introduced by Anthropic in 2024 that defines a universal interface for connecting large language models to external tools, data sources, and services.
Model Pruning
A model compression technique that removes redundant or low-importance parameters from a neural network to reduce size, memory footprint, and inference latency while preserving accuracy.
Model Registry
A model registry is a centralised system that catalogues, versions, and governs trained machine learning models throughout their lifecycle, supporting reproducibility, deployment, and compliance.
Model Serving
Model serving is the discipline of deploying trained machine learning models behind APIs or runtimes so that production applications can request predictions at scale with predictable latency, throughput, and reliability.
Monte Carlo Methods
A broad class of computational algorithms that use repeated random sampling to obtain numerical results, widely used in machine learning for Bayesian inference, reinforcement learning, and uncertainty estimation.
MSC Malaysia
MSC Malaysia, originally the Multimedia Super Corridor, was a Malaysian government initiative established in 1996 to attract technology investment and develop a high-technology economic zone, later rebranded as Malaysia Digital in 2022.
Multi-Agent Systems
Multi-agent systems in AI are architectures in which multiple autonomous AI agents, each with specialised capabilities, collaborate through communication and coordination to complete complex tasks that exceed the capability of any single agent.
Multi-Task Learning
Multi-task learning is a machine learning approach in which a model is trained simultaneously on multiple related tasks, using shared representations to improve generalisation and data efficiency compared to training separate single-task models.
Multimodal AI
Artificial intelligence systems that can process, understand, and generate information across multiple data types simultaneously, including text, images, audio, video, and other modalities.
MyCC and AI Competition Law in Malaysia
An overview of the Malaysia Competition Commission's role in regulating AI-driven digital markets, including its 2025 market review of the digital economy under the Competition Act 2010.
MyDigital Blueprint
The Malaysia Digital Economy Blueprint, known as MyDIGITAL, is Malaysias national strategy to transform the country into a digitally driven, high-income economy by 2030, with AI as a core enabler.
Named Entity Recognition
Named entity recognition (NER) is a natural language processing task that identifies and classifies named entities in text — such as people, organisations, locations, and dates — into predefined categories.
National AI Office Malaysia
The National AI Office (NAIO) is a Malaysian government body established in 2024 to coordinate national artificial intelligence policy, governance, talent development, and adoption across public and private sectors.
National AI Roadmap Malaysia (AI-RMAP 2021–2025)
Malaysia's first comprehensive national strategy for artificial intelligence, adopted in December 2021 and led by MOSTI with MDEC as implementation partner, aiming to position Malaysia among the top 20 nations in AI government readiness by 2025.
National Cyber Security Agency (NACSA)
The Malaysian federal agency responsible for coordinating national-level cyber security policy, protecting Critical National Information Infrastructure, and shaping the country's response to emerging threats including artificial intelligence and AI-enabled attacks.
Natural Language Generation
Natural Language Generation (NLG) is a subfield of artificial intelligence that automatically produces human-readable text from structured data, semantic representations, or other machine-readable inputs.
Natural Language Processing
Natural language processing (NLP) is the subfield of AI concerned with enabling computers to understand, interpret, manipulate, and generate human language in both text and speech form.
Neptune.ai
Neptune.ai was an MLOps experiment tracking and metadata management platform that provided data science teams with tools to log, compare, and reproduce machine learning experiments at scale; the company was acquired by OpenAI in 2025.
Neural Architecture Search
Neural architecture search is the automated design of neural network architectures using search algorithms, reinforcement learning, or gradient-based methods to discover models that meet target accuracy, latency, and size constraints.
Neural Network
A neural network is a computational model inspired by biological brains, composed of interconnected layers of nodes that learn patterns from data through weighted connections.
Neural Scaling Laws
Neural scaling laws are empirical relationships describing how the performance of neural networks improves predictably as a function of model size, dataset size, and compute budget, enabling principled resource allocation for AI training.
Neuro-symbolic AI
Neuro-symbolic AI is a hybrid artificial intelligence paradigm that combines neural network-based learning with symbolic reasoning, integrating the pattern recognition strengths of deep learning with the structured reasoning and interpretability of symbolic methods.
NVIDIA
An American technology company and the world's dominant supplier of graphics processing units (GPUs) for artificial intelligence training and inference, responsible for the CUDA parallel computing platform and a broad ecosystem of AI hardware and software.
NVIDIA Blackwell
NVIDIA Blackwell is a GPU architecture introduced in 2024 for AI training and inference, featuring dual-die GPUs, FP4 precision, and the GB200 Grace Blackwell Superchip and NVL72 rack-scale systems for trillion-parameter models.
Object Detection
Object detection is a computer vision task that involves identifying the location and category of one or more objects within an image or video frame, producing bounding boxes and class labels for each detected instance.
Ollama
Ollama is an open-source runtime that enables developers and researchers to download, run, and manage large language models locally on consumer hardware without cloud API dependencies.
ONNX (Open Neural Network Exchange)
An open standard format for representing machine learning models that enables interoperability between deep learning frameworks, runtimes, and hardware platforms.
OpenAI
An American artificial intelligence research organisation and technology company, founded in 2015, known for developing the GPT series of language models and the ChatGPT conversational AI platform.
OpenVINO
OpenVINO is an open-source toolkit developed by Intel for optimising and deploying deep learning inference across Intel hardware, including CPUs, GPUs, Neural Processing Units, and FPGAs, with broad support for major AI frameworks and model formats.
Optical Character Recognition
A computer vision technology that converts images of typed, handwritten, or printed text into machine-readable digital text, increasingly powered by deep learning and transformer-based vision models.
Overfitting
Overfitting is a modelling error in machine learning where a model learns the training data too closely, including its noise, and consequently performs poorly on new, unseen data.
Parameter-Efficient Fine-Tuning
A family of techniques that adapts a pretrained language or vision model to a downstream task by training only a small fraction of its parameters, dramatically reducing compute, memory, and storage requirements compared to full fine-tuning.
PDPA AI Compliance
PDPA AI compliance refers to the application of Malaysia's Personal Data Protection Act 2010 to artificial intelligence systems, governing how personal data may be collected, processed, and used in AI training, inference, and deployment.
Perplexity AI
Perplexity AI is an American AI company that operates an answer engine combining real-time web search with large language model synthesis, providing cited, conversational responses to user queries.
pgvector
pgvector is an open-source PostgreSQL extension that adds a vector data type and similarity-search operators, allowing embeddings to be stored and queried directly inside a relational database.
Phi (Language Model)
A family of small language models developed by Microsoft Research that demonstrate strong reasoning and instruction-following at parameter counts an order of magnitude smaller than typical frontier models.
Physical AI
Physical AI is artificial intelligence that perceives, reasons about, and acts upon the physical world through embodied systems such as robots, autonomous vehicles, and automated facilities, bridging digital intelligence and real-world action.
Pika Labs
A United States-based artificial intelligence startup founded in 2023 that develops the Pika text-to-video and image-to-video generation models, competing with Runway and OpenAI's Sora in the AI video generation market.
Pinecone
Pinecone is a managed, cloud-native vector database designed for storing high-dimensional embeddings and serving low-latency similarity search for retrieval-augmented AI applications.
Pose Estimation
Pose estimation is the computer vision task of detecting and tracking the position and orientation of human bodies, hands, or objects from images or video, typically by locating keypoints such as joints.
Precision and Recall
Precision and recall are two complementary metrics used to evaluate classification models, measuring respectively the correctness of positive predictions and the completeness with which actual positives are identified.
Predictive Maintenance
Predictive maintenance is the use of sensor data, statistical modelling, and machine learning to forecast equipment failures before they occur, enabling repairs to be scheduled precisely when needed.
Principal Component Analysis
An unsupervised statistical technique that transforms correlated variables into a smaller set of uncorrelated components that preserve as much variance in the original data as possible.
Prompt Caching
Prompt caching is an inference optimisation technique that stores precomputed key-value representations of repeated prompt prefixes, reducing latency and token processing costs for applications with stable system prompts or long shared contexts.
Prompt Engineering
The practice of designing and optimising input instructions given to large language models to elicit accurate, relevant, and well-structured outputs for a given task or application.
Prompt Injection
Prompt injection is a security vulnerability affecting large language model applications in which an attacker embeds adversarial instructions in model inputs to override the system's intended behaviour, bypass safety controls, or exfiltrate sensitive information.
Proximal Policy Optimization
A reinforcement learning algorithm developed by OpenAI that stabilises policy gradient training by constraining the size of policy updates, widely used for fine-tuning large language models through RLHF.
PyTorch
PyTorch is an open-source machine learning framework, originally developed by Meta AI, that provides tensor computation with GPU acceleration and a dynamic computational graph for building and training deep neural networks.
Qdrant
An open-source, Rust-based vector database and similarity search engine designed for high-performance storage and retrieval of high-dimensional embeddings, with support for hybrid search and multitenant deployments.
Quantisation
Quantisation is a model compression technique that reduces the numerical precision of a neural network's weights and activations from high-bit floating-point formats to lower-bit representations, decreasing memory usage and accelerating inference with minimal accuracy loss.
Question Answering
Question answering is the natural language processing task of producing accurate answers to questions posed in natural language, often using information retrieval, reading comprehension, or large language models.
Qwen
Qwen is a family of large language models developed by Alibaba Cloud, ranging from small open-weight dense models to trillion-parameter mixture-of-experts systems, with strong multilingual and reasoning capabilities.
Random Forest
Random forest is an ensemble machine learning algorithm that builds many decision trees on bootstrapped samples and aggregates their predictions to improve accuracy and reduce overfitting.
ReAct (Reasoning and Acting)
ReAct is a prompting framework that interleaves reasoning traces with task actions, letting a language model plan, call external tools, and incorporate observations to solve problems more reliably.
Reasoning Models
Reasoning models are large language models trained to generate extended internal deliberation before producing a final answer, using test-time compute to improve accuracy on complex tasks such as mathematics, coding, and multi-step logic.
Recommendation System
A recommendation system is an information filtering algorithm that predicts and surfaces items — such as products, content, or services — that a particular user is likely to find relevant, based on past behaviour, item attributes, or the preferences of similar users.
Recurrent Neural Network
A recurrent neural network (RNN) is a class of neural network designed for sequential data, where connections between nodes form directed cycles allowing information to persist across time steps.
Regularisation (Machine Learning)
Regularisation is a collection of techniques in machine learning that constrain models during training to reduce overfitting and improve generalisation to unseen data.
Reinforcement Learning
A machine learning paradigm in which an agent learns to make sequential decisions by interacting with an environment and optimising for cumulative reward through trial and error.
Reinforcement Learning from AI Feedback
Reinforcement Learning from AI Feedback (RLAIF) is an alignment technique in which an AI model provides preference labels to train a reward model, replacing or supplementing expensive human annotation used in RLHF.
Reinforcement Learning from Human Feedback
A machine learning technique that trains a reward model from human preference data and uses it to align large language models with human values, safety requirements, and intended behaviour through reinforcement learning.
Reranking
Reranking is a two-stage information retrieval technique in which a fast first-stage retriever generates candidate documents, and a more accurate but computationally expensive model re-scores and reorders them.
Residual Network
A deep convolutional neural network architecture introduced by Microsoft Research in 2015 that uses skip connections to enable training of very deep networks, winning the ImageNet challenge with a top-5 error rate of 3.57%.
Responsible AI
A framework of principles and practices that guide the development and deployment of artificial intelligence systems to ensure they are safe, fair, transparent, accountable, and aligned with human values.
Retrieval-Augmented Generation
A technique that enhances large language model outputs by retrieving relevant documents from an external knowledge base at inference time, grounding responses in up-to-date and domain-specific information.
Reward Modeling
Reward modeling is the process of training a neural network to predict human or AI preferences over model outputs, providing a scalable reward signal for reinforcement learning-based alignment of language models.
Runway ML
Runway is a generative artificial intelligence company that develops video generation and editing models, best known for its Gen-series text-to-video systems used in filmmaking and content creation.
Salesforce Einstein
Salesforce Einstein is an integrated artificial intelligence platform embedded across the Salesforce CRM ecosystem, providing predictive analytics, generative AI, and autonomous agents for sales, service, marketing, and commerce workflows.
SC Malaysia Fintech
The Securities Commission Malaysia (SC) regulates capital market fintech activities including digital assets, robo-advisory, equity crowdfunding, and AI-driven investment services through a framework of licensing, regulatory sandboxes, and guidelines.
Scale AI
An American data labelling, evaluation, and AI infrastructure company that supplies training data and evaluation services to leading AI laboratories, autonomous vehicle developers, and government agencies.
Scikit-learn
Scikit-learn is an open-source Python library for classical machine learning, providing accessible and consistent implementations of classification, regression, clustering, and data-preprocessing algorithms built on NumPy and SciPy.
SEA-LION
SEA-LION (Southeast Asian Languages In One Network) is an open-source family of large language models developed by AI Singapore to serve the languages and cultures of Southeast Asia.
Segment Anything Model
The Segment Anything Model (SAM) is a foundation model from Meta AI for promptable image and video segmentation, able to isolate any object from a click, box, or mask with strong zero-shot generalisation.
Self-Supervised Learning
A machine learning training paradigm in which a model generates its own supervisory signal from unlabelled data by solving pretext tasks, learning rich representations without human-annotated labels.
Semantic Search
Semantic search is a search paradigm that retrieves results based on the meaning and intent of a query rather than exact keyword matches, using vector embeddings to measure conceptual similarity between text.
Sentence Transformers
Sentence Transformers are neural network models that encode sentences, paragraphs, or short documents into fixed-length dense vector embeddings optimised for semantic similarity comparison.
Sentiment Analysis
Sentiment analysis is a natural language processing technique that automatically identifies and classifies the emotional tone of text as positive, negative, or neutral, and is widely used in customer feedback, social media monitoring, and financial analysis.
Sequence-to-Sequence Model
A neural network architecture composed of an encoder that processes an input sequence into a fixed representation and a decoder that generates an output sequence from that representation, forming the foundation for machine translation, summarisation, and dialogue systems.
ServiceNow AI
ServiceNow AI is the artificial intelligence layer embedded in the ServiceNow platform, providing predictive analytics, generative AI, and autonomous agents for IT service management, enterprise workflow automation, and cross-functional business operations.
Shadow Mode
Shadow mode is a machine learning deployment strategy in which a new model processes live production traffic in parallel with the existing model, capturing outputs for evaluation without affecting users or business operations.
Small Language Models
Small language models (SLMs) are compact language models with fewer than around 10 billion parameters, designed for efficient deployment on edge devices, mobile hardware, and resource-constrained environments.
Softmax Function
The softmax function converts a vector of real-valued scores into a probability distribution, and is widely used as the output layer of neural network classifiers and in attention mechanisms.
Sora
Sora is a text-to-video generative AI model developed by OpenAI that produces short, high-fidelity video clips with synchronised audio from natural-language prompts.
Sovereign AI
Sovereign AI is the capacity of a nation to develop, deploy, and govern artificial intelligence using its own infrastructure, data, talent, and models, ensuring strategic autonomy and alignment with domestic laws and values.
Spark
A large language model developed by iFlyTek, a Chinese AI company specialising in speech recognition and natural language processing, notable for its multilingual capabilities covering over 130 languages including Malay and other ASEAN languages.
Sparse Autoencoder
A sparse autoencoder is a type of autoencoder trained with a sparsity constraint that forces most neurons in the hidden layer to be inactive for any given input, producing a disentangled, interpretable feature decomposition.
Speculative Decoding
Speculative decoding is an inference acceleration technique that uses a small draft model to propose multiple candidate tokens that a larger target model then verifies in parallel, achieving 2-4x throughput gains without changing output quality.
Speech Recognition
Speech recognition, or automatic speech recognition (ASR), is the technology that enables computers to identify and transcribe spoken language into text using acoustic models, language models, and deep learning architectures.
Stability AI
A British artificial intelligence company best known for developing and releasing Stable Diffusion, an open-weight text-to-image generative model, and a family of related image, video, audio, and 3D models.
Stable Diffusion
Stable Diffusion is an open-source latent diffusion model developed by Stability AI that generates high-quality images from text prompts, running efficiently on consumer-grade hardware.
Support Vector Machine
A support vector machine (SVM) is a supervised machine learning algorithm that finds the optimal hyperplane separating data points of different classes by maximising the margin between the boundary and the nearest training examples.
Synthetic Data
Synthetic data is artificially generated data that mimics the statistical properties of real datasets, created using generative AI or simulations to train machine learning models without exposing sensitive personal information.
TechCity KL
TechCity KL is a technology and innovation district initiative in Kuala Lumpur designed to attract technology companies, AI startups, and multinational regional headquarters to a dedicated urban precinct within the Malaysian capital.
Tensor Processing Unit
A tensor processing unit (TPU) is a custom application-specific integrated circuit developed by Google for accelerating machine learning workloads, particularly neural network training and inference.
TensorFlow
TensorFlow is an open-source machine learning platform developed by Google that supports the full lifecycle of building, training, and deploying models across servers, mobile devices, browsers, and edge hardware.
TensorFlow Lite
TensorFlow Lite is an open-source deep learning framework from Google for running optimised machine learning models on mobile phones, microcontrollers, and other edge devices.
Text Summarisation
Text summarisation is the natural language processing task of producing a shorter version of a document that preserves its key information, using extractive or abstractive techniques.
Text-to-Speech
Text-to-speech is the technology that converts written text into synthesised spoken audio using rule-based, concatenative, or neural network methods.
Time Series Forecasting
Time series forecasting is the application of statistical and machine learning methods to predict future values of a sequence of observations indexed in time, such as sales, demand, electricity load, or financial prices.
TinyML
TinyML is a field of machine learning focused on running machine learning models on microcontrollers and other resource-constrained edge devices that typically operate with milliwatts of power and kilobytes of memory.
Token
A token is the smallest unit of text processed by a large language model, typically representing a word, subword, or character used as the fundamental input and output element during inference.
Tokenisation
Tokenisation is the process of breaking text into discrete units called tokens — which may represent words, subwords, characters, or symbols — that serve as the fundamental input units for language models and other natural language processing systems.
Tool Use
Tool use in AI refers to the capability of language models to invoke external functions, APIs, or services to retrieve information, perform actions, or extend their abilities beyond text generation.
Transfer Learning
Transfer learning is a machine learning technique in which a model pre-trained on one task or dataset is adapted for a different but related task, enabling high performance with significantly less data and compute than training from scratch.
Transformer Architecture
A neural network architecture introduced in 2017 that uses self-attention mechanisms to process sequential data in parallel, forming the foundation of modern large language models and multimodal AI systems.
Tree of Thoughts
Tree of Thoughts (ToT) is a prompting framework that lets large language models explore and evaluate multiple intermediate reasoning paths in a branching search, improving performance on tasks requiring planning.
Variational Autoencoder
A variational autoencoder is a generative neural network that learns a probabilistic latent representation of data, enabling smooth sampling and reconstruction of new examples.
Vector Database
A specialised database system that stores data as high-dimensional numerical vectors and enables fast approximate nearest-neighbour search, forming the retrieval backbone of semantic search and RAG systems.
Vibe Coding
Vibe coding is an AI-assisted software development practice in which a developer describes intent in natural language and a large language model generates the code, with the human guiding and testing rather than writing it directly.
Vision Transformer
The Vision Transformer (ViT) is a deep learning model that applies the transformer architecture originally designed for NLP directly to sequences of image patches, achieving state-of-the-art results on visual recognition tasks.
Vision-Language Model
A multimodal AI system that jointly processes and generates information from both images and text, extending large language models with visual perception capabilities through cross-modal alignment.
vLLM
vLLM is an open-source library for fast and memory-efficient large language model inference and serving, built around the PagedAttention algorithm for optimised GPU memory management.
Weaviate
An open-source, cloud-native vector database that combines vector similarity search with structured filtering, GraphQL APIs, and built-in vectorisation for AI applications.
Weights and Biases
Weights and Biases (W&B) is a machine learning developer platform for experiment tracking, model versioning, dataset management, and collaborative model evaluation used by over 200,000 ML practitioners worldwide.
Whisper
Whisper is an open-source automatic speech recognition system developed by OpenAI, trained on 680,000 hours of multilingual audio data and capable of transcription, translation, and language identification across nearly 100 languages.
Word2Vec
A neural network-based algorithm developed by Google in 2013 that learns dense vector representations of words from large text corpora, capturing semantic and syntactic relationships through distributional similarity.
World Models
World models are AI systems that build internal representations of how the environment works, enabling machines to simulate, plan, and reason about future states without requiring direct experience.
XGBoost
XGBoost (Extreme Gradient Boosting) is an open-source machine learning library that provides a fast, regularised gradient boosting framework, widely used for classification, regression, and ranking on tabular data.
Yi
A family of open-source bilingual large language models developed by 01.AI, the Beijing-based AI startup founded by Kai-Fu Lee, achieving competitive performance against Llama 2 and Falcon with strong Chinese and English bilingual capability.
YOLO (You Only Look Once)
YOLO is a family of real-time object detection models that frame detection as a single regression problem, predicting bounding boxes and class probabilities directly from an image in one network pass.
Zero-Shot Learning
Zero-shot learning is a machine learning paradigm in which a model makes accurate predictions on categories it has never seen during training by leveraging semantic descriptions or attribute representations.