large language model judge calibration
LLM judge calibration
Large language model judge calibration adjusts and validates a model-based evaluator against trusted human-reviewed examples so its scores reflect the intended quality standard.
AI Terminology
Search clear AI terminology definitions, acronyms, related concepts and the reviewed videos that explain them.
Showing 481–500 of 705 terms
Clear filterslarge language model text file
llms.txtA large language model text file is a concise website file, commonly named llms.txt, that points AI systems to authoritative descriptions and documentation for the site or project.
latency-aware artificial intelligence model routing
Latency-aware AI model routing chooses models using response-time requirements and current or expected service speed.
Lean proof assistant
Lean is a proof assistant and functional programming language used to define mathematical objects, write programs and create machine-checked proofs.
linear attention
Linear attention reformulates attention so its computation grows roughly linearly with sequence length rather than quadratically like standard full attention.
linter
AI linterAn AI linter is an automated reviewer that uses a model to identify code issues that are difficult to express as fixed syntactic rules.
literacy
AI literacyAI literacy is the foundational ability to understand what AI can and cannot do, how it affects decisions and how to use it critically and responsibly.
local artificial intelligence runtime
local AI runtimeA local AI runtime is software that loads and executes AI models on hardware controlled by the user.
log compaction
WAL compactionLog compaction reduces an append-only log by preserving the state or records needed for recovery while removing superseded history.
machine-checked proof
A machine-checked proof is a formal proof whose logical steps are represented in a form that a small verification program can independently validate.
machine-readable documentation
Machine-readable documentation is authoritative product or technical guidance formatted and organized so software can retrieve, parse, and use it reliably without depending on visual page interpretation.
machine-readable product information
Machine-readable product information is current product data expressed in a structured format that software can reliably parse, compare, validate, and use without inferring meaning from presentation alone.
managed graphics processing unit cluster
managed GPU clusterA managed graphics processing unit cluster is accelerator infrastructure operated and maintained by a provider for a customer's AI workloads.
Markdown
Markdown is a lightweight plain-text markup format that uses simple punctuation to represent headings, links, lists, code, emphasis, and other document structure.
market midpoint
A market midpoint is the price halfway between the best available bid and best available ask.
message queue
A message queue stores work or events until a consumer can process them, separating producers from consumers in time and capacity.
minimum viable artificial intelligence solution
minimum viable AI solutionsmallest responsible AI solutionA minimum viable AI solution is the smallest bounded system that can test whether AI improves a real outcome responsibly.
mobile artificial intelligence workflow
A mobile AI workflow lets a user start, continue, review or control AI-assisted work from a phone or tablet.
model architecture
An AI model architecture is the structural design that defines a model's components, connections and flow of information.
model bias
AI model biasAI model bias is a systematic tendency in a model's outputs that reflects imbalanced data, design choices, objectives or deployment context.
Knowledge graph
Explore AI concepts
Follow general terms into specialised sub-terms. Select any node to open its definition.