test-time reasoning
Test-time reasoning lets an AI model spend additional inference work analyzing a specific request before it produces or finalizes an answer.
AI Terminology
Search clear AI terminology definitions, acronyms, related concepts and the reviewed videos that explain them.
Showing 1021–1040 of 1091 terms
Clear filterstext style transfer
Text style transfer rewrites language to change its tone or stylistic properties while preserving the intended meaning.
text-to-speech
Text-to-speech converts written text into spoken audio by predicting pronunciation, timing, prosody and a waveform or intermediate audio representation.
tiered AI model routing
Tiered AI model routing assigns requests to capability and cost tiers, using different models for routine and difficult work.
time to first token
TTFTTime to first token is the delay between sending a model request and receiving the first generated output token.
time-series forecasting
Time-series forecasting predicts future values from observations recorded in time order.
token accounting
AI token accounting records the input, output and related token usage associated with model requests.
tokenizer
A tokenizer converts text or other model input into token identifiers that an AI model can process.
tool integration
Tool integration connects an AI system to an external application, service, or data source that it can use during a task.
tool output spoofing
Tool output spoofing makes fabricated or altered tool results appear genuine to an AI evaluator, agent, or downstream system.
tournament sampling
Tournament sampling draws several candidates from a probability distribution and selects a winner through repeated pairwise decisions.
trading agent
A trading agent is an automated system that observes markets and places or manages trades according to a strategy.
training data attribution
Training data attribution estimates which training examples most influenced a particular AI model output or behavior.
transcript summarization
Transcript summarization uses a person or AI system to condense spoken-content text into its most useful arguments, decisions, actions or sections.
transfer learning
Transfer learning reuses knowledge learned from earlier data or tasks to improve learning on a new task.
transferable compute capacity contract
A transferable compute capacity contract reserves computing resources while allowing the holder to assign or resell some contractual rights under defined conditions.
transparency
AI transparency makes an AI system's purpose, limits, evidence, operation and responsible parties understandable to the people who use or are affected by it.
trusted kernel
A trusted kernel is the small core of a proof system that independently checks whether formal proof terms are valid under the system's logical rules.
typed event
A typed event is a named system event whose payload and meaning follow a defined schema or contract.
uncertainty quantification
UQAI uncertainty quantification estimates and represents how uncertain a model is about predictions, hidden quantities, data, or possible outcomes.