Difficulty: Intermediate

412 terms

Human-in-the-Loop Intermediate

System design where humans validate or guide model outputs, especially for high-stakes decisions.

Hyperparameters Intermediate

Configuration choices not learned directly (or not typically learned) that govern training or architecture.

Image Classification Intermediate

Assigning category labels to images.

Incident Response Intermediate

Process for managing AI failures.

Inductive Bias Intermediate

Built-in assumptions guiding learning efficiency and generalization.

Inference Cost Intermediate

Cost to run models in production.

Inference Pipeline Intermediate

Model execution path in production.

Information Gain Intermediate

Reduction in uncertainty achieved by observing a variable; used in decision trees and active learning.

Informed Consent Intermediate

Patient agreement to AI-assisted care.

Instance Segmentation Intermediate

Pixel-level separation of individual object instances.

Inter-Annotator Agreement Intermediate

Measure of consistency across labelers; low agreement indicates ambiguous tasks or poor guidelines.

Interpretability Intermediate

Studying internal mechanisms or input influence on outputs (e.g., saliency maps, SHAP, attention analysis).

ISO/IEC 23894 Intermediate

International AI risk standard.

Kalman Filter Intermediate

Optimal estimator for linear dynamic systems.

Key-Value Cache Intermediate

Stores past attention states to speed up autoregressive decoding.

Kill Switch Intermediate

Mechanism to disable AI system.

KL Divergence Intermediate

Measures how one probability distribution diverges from another.

Knowledge Graph Intermediate

Structured graph encoding facts as entity–relation–entity triples.

Lagrangian Intermediate

Converts constrained problem to unconstrained form.

Language Model Intermediate

A model that assigns probabilities to sequences of tokens; often trained by next-token prediction.

Large Language Model Intermediate

A high-capacity language model trained on massive corpora, exhibiting broad generalization and emergent behaviors.

Latency Intermediate

Time from request to response; critical for real-time inference and UX.

Latency SLA Intermediate

Guaranteed response times.

Latent Space Intermediate

The internal space where learned representations live; operations here often correlate with semantics or generative factors.

Learning Rate Intermediate

Controls the size of parameter updates; too high diverges, too low trains slowly or gets stuck.

Learning Rate Schedule Intermediate

Adjusting learning rate over training to improve convergence.

Legal AI Intermediate

AI supporting legal research, drafting, and analysis.

Legal Hold Intermediate

Requirement to preserve relevant data.

LIME Intermediate

Local surrogate explanation method approximating model behavior near a specific input.

Line Search Intermediate

Choosing step size along gradient direction.