inference benchmark
An AI inference benchmark measures model-serving performance under a defined workload, hardware configuration, and software stack.
AI Terminology
Search clear AI terminology definitions, acronyms, related concepts and the reviewed videos that explain them.
Showing 381–400 of 1136 terms
Clear filtersintegrated development environment
IDEAn integrated development environment is a software workspace that combines code editing with development tools such as navigation, debugging, testing and extensions.
knowledge graph
A knowledge graph represents entities, concepts and facts as connected nodes and relationships that software can query and traverse.
Lean proof assistant
Lean is a proof assistant and functional programming language used to define mathematical objects, write programs and create machine-checked proofs.
linter
An AI linter is an automated reviewer that uses a model to identify code issues that are difficult to express as fixed syntactic rules.
machine-speed cyberattack
A machine-speed cyberattack uses automation to perform malicious activity faster or across more targets than a person could manage manually.
Markdown
Markdown is a lightweight plain-text notation for adding headings, links, lists, code, emphasis, and other document structure.
message-based AI agent interface
A message-based AI agent interface lets software exchange structured requests, context, actions, and results with an agent through a defined messaging contract.
model-agnostic platform
A model-agnostic platform supports interchangeable AI models or providers behind one product experience instead of depending on one exclusive model stack.
multi-robot task allocation
MRTAMulti-robot task allocation assigns parts of a shared objective to robots according to their capabilities, locations and current workload.
object storage
Object storage keeps data as independently addressed objects with payloads and metadata in a flat, scalable namespace.
offensive AI
Offensive AI uses AI to assist or automate actions intended to discover, exploit, disrupt, manipulate, or compromise a target.
organizational AI coordinator
An organizational AI coordinator is an agent or system that routes and coordinates work across teams, tools, and organizational boundaries.
payment settlement
Payment settlement is the process by which a payment reaches an accepted final state and the recipient can rely on the transferred value.
post-training
Post-training adapts a pretrained AI model to improve its behavior, capabilities, or alignment after the main training stage.
prompt optimization
Prompt optimization searches for instructions or examples that improve an AI system's measured performance on a defined task without necessarily changing its model weights.
research iteration cycle
A research iteration cycle is the repeated process of forming an idea, running an experiment, evaluating evidence and choosing the next step.
reversible AI agent action
A reversible AI agent action is designed with a validated procedure that can safely undo or compensate for its effects.
robot reasoning and control stack
A robot reasoning and control stack is the layered software that interprets goals, plans actions and controls a robot's physical execution.
robot-to-robot communication
Robot-to-robot communication is the exchange of task, state or environment information directly or through a shared coordination service.