AI dictionary
Browse AI terms A–Z
Search or browse the full dictionary below, or explore topic hubs when you want to learn an area step by step.
636 terms
Search & index
All terms A–Z
Filter by name, or jump to a letter using the bar below.
636 terms · A–Z
A
- A/B testing
- ablation
- accelerator chip
- accuracy
- act
- action
- action space
- activation function
- active learning
- AdaGrad
- adaptation
- Agent
- agent orchestration
- agentic
- agentic loop
- agentic workflow
- agglomerative clustering
- AI slop
- anomaly detection
- area under the PR curve
- area under the ROC curve
- artificial general intelligence
- Artificial Intelligence
- Attention
- attribute
- attribute sampling
- AUC
- augmented reality
- auto-regressive model
- autoencoder
- automatic evaluation
- automation bias
- AutoML
- autonomous agent
- autorater evaluation
- auxiliary loss
- average precision at k
- axis-aligned condition
B
- Backpropagation
- bag of words
- bagging
- base model
- baseline
- batch
- batch inference
- batch normalization
- batch size
- Bayesian neural network
- Bayesian optimization
- Bellman equation
- BERT
- bias
- bias (math) or bias term
- bidirectional
- bidirectional language model
- bigram
- binary classification
- binary condition
- binning
- black box model
- BLEU
- BLEURT
- boosting
- bounding box
- broadcasting
- bucketing
C
- calibration layer
- candidate generation
- candidate sampling
- categorical data
- causal language model
- centroid
- centroid-based clustering
- Chain-of-Thought Prompting
- Character N-gram F-score
- chat
- checkpoint
- citation precision
- citation recall
- class
- class-balanced dataset
- class-imbalanced dataset
- Classification
- classification model
- classification threshold
- classifier
- clipping
- Clustering
- co-adaptation
- co-training
- collaborative filtering
- compact model
- compute
- concept drift
- condition
- Confabulation
- configuration
- confirmation bias
- confusion matrix
- constituency parsing
- Context Window
- contextualized language embedding
- continuous feature
- convenience sampling
- convergence
- conversational coding
- convex function
- convex optimization
- convex set
- convolution
- convolutional filter
- convolutional layer
- convolutional neural network
- convolutional operation
- cost
- counterfactual fairness
- coverage bias
- crash blossom
- critic
- cross-entropy
- cross-validation
- cumulative distribution function
D
- data analysis
- Data Augmentation
- data parallelism
- data set or dataset
- decision boundary
- decision forest
- decision threshold
- decision tree
- decoder
- deep model
- deep neural network
- Deep Q-Network
- demographic parity
- denoising
- dense feature
- dense layer
- depth
- depthwise separable convolutional neural network
- derived label
- deterministic
- differential privacy
- dimension reduction
- dimensions
- direct prompting
- discrete feature
- discriminative model
- discriminator
- disparate impact
- disparate treatment
- distillation
- distribution
- divisive clustering
- downsampling
- dropout regularization
- dynamic
- dynamic model
E
- early stopping
- earth mover's distance
- edit distance
- Einsum notation
- Embedding
- embedding layer
- embedding space
- embedding vector
- emergent behavior
- empirical cumulative distribution function
- empirical risk minimization
- encoder
- ensemble
- entropy
- environment
- environment grounding
- episode
- episodic memory
- Epoch
- epsilon greedy policy
- equality of opportunity
- equalized odds
- evals
- evaluation
- evaluator agent
- exact match
- example
- experience replay
- experimenter's bias
- exploding gradient problem
F
- F1
- factuality
- fairness constraint
- fairness metric
- false negative
- false negative rate
- false positive
- false positive rate
- fast decay
- feature
- feature cross
- feature engineering
- feature extraction
- feature importances
- feature set
- feature spec
- feature vector
- featurization
- federated learning
- feedback
- feedback loop
- feedforward neural network
- few-shot learning
- few-shot prompting
- Fine-tuning
- forget gate
- foundation model
- fraction of successes
- full softmax
- fully connected layer
- function transformation
G
- Gemini
- GenAI or genAI
- generalization
- generalization curve
- generalized linear model
- generated text
- generative adversarial network
- generative agents
- generative AI
- generative model
- generator
- gini impurity
- golden dataset
- golden response
- GPT
- gradient
- gradient accumulation
- gradient boosted (decision) trees
- gradient boosting
- gradient clipping
- Gradient Descent
- graph
- greedy policy
- ground truth
- groundedness
- grounding
- group attribution bias
- guardrails
H
I
- i.i.d.
- image recognition
- imbalanced dataset
- implicit bias
- imputation
- in-context learning
- in-group bias
- in-set condition
- incompatibility of fairness metrics
- independently and identically distributed
- individual fairness
- Inference
- inference path
- information gain
- input generator
- input layer
- instance
- instruction tuning
- inter-rater agreement
- interpretability
- intersection over union
- item matrix
- items
- iteration
K
L
- L0 regularization
- L1 loss
- L1 regularization
- L2 loss
- L2 regularization
- label
- label leakage
- labeled example
- lambda
- landmarks
- language model
- large language model
- latency
- latent space
- layer
- leaf
- Learning Interpretability Tool
- learning rate
- least squares regression
- least-to-most prompting
- Levenshtein Distance
- linear
- linear model
- linear regression
- LLM
- LLM evaluations
- Log Loss
- log-odds
- logistic regression
- logits
- Long Short-Term Memory
- LoRA
- loss
- loss aggregator
- loss curve
- loss function
- loss surface
- lost-in-the-middle effect
- Low-Rank Adaptability
M
- Machine Learning
- machine translation
- majority class
- manager agent
- Markov decision process
- Markov property
- masked language model
- math-pass@k
- matrix factorization
- MBPP
- MCP
- Mean Absolute Error
- mean average precision at k
- Mean Squared Error
- meta-learning
- metric
- mini-batch
- mini-batch stochastic gradient descent
- minimax loss
- minority class
- mixture of experts
- MMIT
- modality
- model
- model capacity
- model cascading
- model parallelism
- model router
- model training
- Momentum
- multi-agent collaboration
- multi-class classification
- multi-class logistic regression
- multi-head self-attention
- multimodal instruction-tuned
- multimodal model
- multinomial classification
- multinomial regression
- multitask
N
- N-gram
- NaN trap
- natural language processing
- natural language understanding
- negative class
- negative sampling
- Neural Architecture Search
- Neural Network
- neuron
- no one right answer
- node
- noise
- non-binary condition
- non-response bias
- nondeterministic
- nonlinear
- nonstationarity
- normalization
- novelty detection
- numerical data
O
- objective
- objective function
- oblique condition
- observe
- offline
- offline inference
- one right answer
- one-hot encoding
- one-shot learning
- one-shot prompting
- one-vs.-all
- online
- online inference
- optimizer
- out-group homogeneity bias
- out-of-bag evaluation
- outlier detection
- outliers
- output layer
- Overfitting
- oversampling
P
- packed data
- parameter
- parameter update
- parameter-efficient tuning
- partial derivative
- participation bias
- partitioning strategy
- pass at k
- perceptron
- performance
- permutation variable importances
- perplexity
- pipeline
- pipelining
- plan-and-solve
- plugin
- policy
- pooling
- positional encoding
- positive class
- post-processing
- post-trained model
- PR AUC
- pre-trained model
- pre-training
- precision
- precision at k
- precision-recall curve
- prediction
- prediction bias
- predictive ML
- predictive parity
- predictive rate parity
- preprocessing
- prior belief
- probabilistic
- probabilistic regression model
- probability density function
- procedural memory
- Prompt
- prompt chaining
- prompt design
- prompt engineering
- prompt set
- prompt tuning
- prompt-based learning
- provenance
- proxy
- proxy labels
- pure function
Q
R
- R-squared
- RAG
- random forest
- random policy
- rank
- ranking
- rater
- re-ranking
- reason
- recall
- recall at k
- recommendation system
- Rectified Linear Unit
- recurrent neural network
- reference text
- reflection
- regression model
- regularization
- regularization rate
- reinforcement learning
- Reinforcement Learning from Human Feedback
- ReLU
- replay buffer
- reporting bias
- representation
- response
- response set
- retrieval-augmented generation
- return
- reward
- ridge regularization
- RNN
- ROC (receiver operating characteristic) Curve
- role prompting
- root
- root directory
- Root Mean Squared Error
- rotational invariance
- ROUGE
- ROUGE-L
- ROUGE-N
- ROUGE-S
- router agent
S
- sampling bias
- sampling with replacement
- SavedModel
- Saver
- scalar
- scaling
- scikit-learn
- scoring
- selection bias
- self-attention
- self-correction
- self-supervised learning
- self-training
- semantic memory
- semi-supervised learning
- sensitive attribute
- sentiment analysis
- sequence model
- sequence-to-sequence task
- serving
- shape
- shard
- shrinkage
- side-by-side evaluation
- sigmoid function
- similarity measure
- single program / multiple data
- size invariance
- sketching
- skip-gram
- soft prompt tuning
- softmax
- sparse feature
- sparse representation
- sparse vector
- sparsity
- spatial pooling
- specificational coding
- split
- splitter
- squared hinge loss
- squared loss
- staged training
- state
- state machine agent
- state-action value function
- static
- static inference
- stationarity
- step
- step size
- stochastic gradient descent
- stride
- structural risk minimization
- sub-agent
- subsampling
- subword token
- summary
- Supervised Learning
- supervised machine learning
- synthetic feature
T
- tabular Q-learning
- target
- target network
- task
- task decomposition
- temperature
- temporal data
- Tensor
- Tensor rank
- Tensor shape
- Tensor size
- termination condition
- test
- test loss
- test set
- text span
- threshold
- time series analysis
- timestep
- Token
- tokenizer
- top-k accuracy
- toxicity
- training
- training loss
- training set
- training-serving skew
- trajectory
- transfer learning
- Transformer
- translational invariance
- tree-of-thought prompting
- trigram
- true negative
- true positive
- true positive rate
U
V
W
X
Z
No terms match your filter.
Browse by topic
Topic hubs
Explore terminology organised by subject — useful when you are learning an area rather than looking up one word.
AI Basics
Fundamental concepts for understanding artificial intelligence.
27 terms
Machine Learning
Algorithms and methods that enable systems to learn from data.
391 terms
Deep Learning
Neural networks with many layers that learn hierarchical representations.
94 terms
Natural Language Processing
AI techniques for understanding and generating human language.
45 terms
Computer Vision
Teaching machines to interpret and understand visual information.
17 terms
Large Language Models
AI models trained on vast text data to understand and generate language.
151 terms
AI Agents
Autonomous systems that perceive, decide, and act to achieve goals.
40 terms
Prompt Engineering
The craft of designing effective inputs for AI models.
49 terms
Reinforcement Learning
Learning through trial and error with rewards and penalties.
25 terms
AI Hardware
Specialised chips and infrastructure for running AI workloads.
8 terms
AI Companies
Organisations building and deploying artificial intelligence.
2 terms
AI Models
Specific AI systems, architectures, and model families.
10 terms
Acronyms
Common abbreviations used throughout the AI field.
15 terms
Ethics & Safety
Responsible development and deployment of AI systems.
49 terms
Mathematics
Mathematical foundations underlying AI and machine learning.
90 terms
Programming Frameworks
Software libraries and tools for building AI applications.
25 terms