Glossary
Key terms in quantum computing and artificial intelligence, from qubits to transformers.
A
- Algorithm (Quantum)
- A computational procedure designed to run on a quantum computer, exploiting superposition and interference to solve certain problems faster than any known classical algorithm.
- Ancilla Qubit
- An auxiliary qubit used to assist a quantum computation — for example, to detect errors or mediate gate operations — without being part of the final output.
- Attention Mechanism
- A component of neural networks that allows the model to weight the importance of different parts of the input when producing an output. The core of the Transformer architecture.
B
- Bloch Sphere
- A unit sphere used to geometrically represent the state of a single qubit. Points on the surface correspond to pure states; the north pole is |0⟩, the south pole is |1⟩.
C
- Coherence
- The property of a quantum system maintaining a stable phase relationship between its quantum states. Coherence is required for quantum interference and is destroyed by decoherence.
- CNOT Gate
- Controlled-NOT gate. A two-qubit gate that flips the target qubit if and only if the control qubit is |1⟩. Fundamental for creating entanglement.
D
- Decoherence
- The process by which a quantum system loses its quantum properties through unwanted interaction with its environment, causing quantum superpositions to collapse into classical states.
- Deep Learning
- A class of machine learning using neural networks with many layers (deep architectures) to learn hierarchical data representations. Powers modern image recognition, language models, and more.
E
- Embedding
- A learned mapping from discrete objects — such as words or tokens — into a continuous high-dimensional vector space, where similar items are placed near each other.
- Entanglement (Quantum)
- A quantum phenomenon where two or more qubits become correlated such that measuring one instantly determines the state of the other, regardless of distance. A key resource for quantum computing and cryptography.
- Error Correction (Quantum)
- Techniques that protect quantum information from decoherence and gate errors by encoding logical qubits redundantly across multiple physical qubits, detecting and correcting errors without collapsing the state.
F
- Fidelity
- A measure from 0 to 1 of how closely a quantum state or operation matches an ideal target. A fidelity of 1 means perfect agreement; 0 means completely orthogonal.
- Fine-tuning
- The process of continuing to train a pre-trained model on task-specific data to adapt its behaviour for a particular application, using a much smaller dataset than the original pre-training.
- Foundation Model
- A large AI model trained on broad, general data (text, images, code) at scale, designed to be adapted for many downstream tasks via fine-tuning or prompting.
G
- Generative AI
- AI systems capable of producing new content — text, images, audio, code — by learning the statistical patterns of their training data. Powered by models such as diffusion models and LLMs.
- Gradient Descent
- An iterative optimization algorithm that updates model parameters in the direction of the steepest decrease in a loss function. The backbone of neural network training.
- Grover's Algorithm
- A quantum search algorithm that finds a marked item in an unsorted list of N items in O(√N) steps — a quadratic speedup over classical brute-force search.
H
- Hadamard Gate
- A single-qubit quantum gate that maps |0⟩ to an equal superposition of |0⟩ and |1⟩, and vice versa. Used to initialize qubits into superposition at the start of quantum algorithms.
- Hallucination
- The tendency of a large language model to generate text that is plausible-sounding but factually incorrect or entirely fabricated, without any explicit indication that it is doing so.
- Hamiltonian
- The operator in quantum mechanics corresponding to the total energy of a system. Simulating a molecule's Hamiltonian on a quantum computer is one of the most promising near-term applications.
- Hybrid Quantum-Classical Computing
- A computational paradigm that uses quantum processors for tasks that benefit from quantum parallelism, while classical processors handle control logic, optimization, and the rest of the computation.
I
- Interference (Quantum)
- A quantum effect where probability amplitudes add constructively or destructively. Quantum algorithms are designed so correct answers interfere constructively and wrong answers cancel out.
L
- Large Language Model (LLM)
- A neural network with billions of parameters trained on massive text corpora to predict and generate language. Forms the backbone of modern AI assistants and coding tools.
- Logical Qubit
- A fault-tolerant qubit encoded across many physical qubits using quantum error correction, intended to be robust enough for practical quantum computation.
M
- Measurement
- The act of observing a quantum system, which irreversibly collapses it from a superposition into one definite classical outcome with a probability determined by the quantum state.
N
- Neural Network
- A machine learning model loosely inspired by biological neurons, composed of layers of parameterized linear transformations and nonlinear activations, trained by gradient descent.
- NISQ
- Noisy Intermediate-Scale Quantum. The current era of quantum hardware, characterized by devices with 50–1000 qubits that are too noisy for full error correction but large enough for some practical experiments.
- No-Cloning Theorem
- A fundamental result in quantum information theory: it is impossible to create an identical copy of an arbitrary unknown quantum state. This underpins the security of quantum cryptography.
O
- Oracle
- A black-box subroutine in a quantum algorithm that encodes information about the problem being solved, typically by flipping the phase of marked states. Used in Grover's and other algorithms.
P
- Physical Qubit
- An actual quantum mechanical system — such as a superconducting circuit, trapped ion, or photon — used to realize a qubit in hardware.
- Prompt Engineering
- The practice of crafting and structuring input prompts to guide a large language model toward producing desired outputs, without changing the model's weights.
Q
- QAOA
- Quantum Approximate Optimization Algorithm. A variational hybrid algorithm designed to find approximate solutions to combinatorial optimization problems using a parameterized quantum circuit.
- QML (Quantum Machine Learning)
- A research field exploring how quantum computers can accelerate or enhance machine learning tasks, including training, inference, and data representation.
- Quantum Advantage
- The point at which a quantum computer performs a task significantly faster, cheaper, or more accurately than the best available classical algorithm. A central goal of quantum computing research.
- Quantum Circuit
- A model of quantum computation represented as a sequence of quantum gates applied to a set of qubits, analogous to a classical digital circuit.
- Quantum Gate
- The quantum analogue of a classical logic gate — a reversible, unitary operation applied to one or more qubits to transform their state.
- Quantum Key Distribution (QKD)
- A cryptographic protocol that uses quantum mechanics to securely distribute encryption keys between two parties, with security guaranteed by the laws of physics rather than computational hardness.
- Quantum Neural Network (QNN)
- A machine learning model using parameterized quantum circuits as trainable layers, analogous to layers in a classical neural network.
- Quantum Simulation
- Using a quantum computer to model and study quantum systems — such as molecules or materials — that are intractable to simulate classically. One of the most promising near-term applications.
- Quantum Speedup
- The computational advantage a quantum algorithm offers over the best known classical algorithm for the same problem, often expressed as a reduction in time or query complexity.
- Quantum Teleportation
- A protocol that transfers an arbitrary quantum state from one location to another using a pair of entangled qubits and two classical bits of communication. No physical matter is transported.
- Quantum Volume
- A hardware-agnostic benchmark metric developed by IBM that quantifies the overall capability of a quantum computer, accounting for qubit count, connectivity, gate fidelity, and error rates.
- Qubit
- The fundamental unit of quantum information. Unlike a classical bit (0 or 1), a qubit can exist in a superposition of |0⟩ and |1⟩ simultaneously, enabling quantum parallelism.
R
- RAG (Retrieval-Augmented Generation)
- An AI technique that grounds a language model's responses by retrieving relevant documents from an external knowledge base at inference time, reducing hallucination.
- Reinforcement Learning
- A machine learning paradigm where an agent learns to make sequential decisions by receiving scalar rewards or penalties from an environment, optimizing for long-term cumulative reward.
S
- Shor's Algorithm
- A quantum algorithm for factoring integers in polynomial time, exponentially faster than the best known classical algorithms. Its practicality would break RSA and other widely used cryptographic systems.
- Superposition
- The quantum mechanical principle that a quantum system can exist in a combination of multiple states simultaneously. When measured, it collapses to a single definite outcome with probabilities given by the amplitudes.
- Surface Code
- A leading quantum error correction code that encodes logical qubits on a 2D lattice of physical qubits, detecting errors by measuring neighboring stabilizer operators without disturbing the encoded information.
T
- T1 Time
- The longitudinal relaxation time of a qubit — how long it takes for an excited |1⟩ state to spontaneously decay to the ground |0⟩ state. A key figure of merit for quantum hardware.
- T2 Time
- The transverse relaxation (dephasing) time of a qubit — how long a qubit maintains phase coherence. T2 ≤ 2T1 always, and longer T2 means more useful quantum computation time.
- Transfer Learning
- A machine learning technique where a model pre-trained on one task is reused as the starting point for training on a different but related task, reducing the need for labelled data.
- Transformer
- A neural network architecture introduced in "Attention Is All You Need" (2017) that relies entirely on self-attention mechanisms. The dominant architecture for large language models and many vision models.
U
- Unitary Operation
- A linear transformation on a quantum state that preserves the total probability (norm). All quantum gates are unitary, making quantum computation inherently reversible.
V
- VQE (Variational Quantum Eigensolver)
- A hybrid quantum-classical algorithm that uses a parameterized quantum circuit to approximate the ground-state energy of a Hamiltonian, with classical optimization updating the circuit parameters.
- Vector Database
- A database system specialized for storing, indexing, and querying high-dimensional vector embeddings. Widely used in RAG pipelines and semantic search.
W
- Wave Function
- A mathematical object describing the complete quantum state of a system. Its squared magnitude gives the probability distribution of measurement outcomes.
- Wave Function Collapse
- The sudden, irreversible transition of a quantum system from a superposition of states to a single definite state upon measurement.