safety culture
Safety culture is the set of shared practices, incentives and expectations that shapes how an organization identifies and responds to risks.
Search clear AI terminology definitions, acronyms, related concepts and the reviewed videos that explain them.
Showing 1261–1280 of 1477 terms
Clear filtersSafety culture is the set of shared practices, incentives and expectations that shapes how an organization identifies and responds to risks.
Sample size is the number of observations used in an analysis, which affects how precisely the evidence can distinguish a real pattern from random variation.
A satisfiability modulo theories solver determines whether logical constraints are simultaneously satisfiable while reasoning about theories such as arithmetic, arrays and bit vectors.
A scalable cyberattack can be repeated or automated across many targets with relatively little additional effort for each new target.
The scaling hypothesis holds that increasing training resources can produce more capable models.
AI scenario forecasting describes a coherent possible path for future AI capabilities, infrastructure, economics, deployment, and societal effects.
A scheduled agent task is an AI workflow configured to begin automatically at a chosen time or repeating interval.
A scheduled AI agent routine is a recurring agent workflow triggered automatically at specified times or intervals.
A scheduled AI deliverable is a document, report or other artifact generated automatically at a defined time or interval.
A screening hit rate is the proportion of tested candidates that meet a predefined activity or performance threshold in a screening experiment.
Secret-key language model watermark detection uses a protected key to reconstruct the token-selection rule and test whether text follows its hidden pattern.
A secure default is a setting or behavior that starts in a protective state without requiring extra action from the user.
AI security scanning uses AI to inspect software, configurations or activity for vulnerabilities and suspicious patterns.
A seed round is an early funding round that gives a young company capital to build its team, product and initial evidence of demand.
Self-report bias occurs when people's accounts of their own behavior or results differ systematically from what happened.
A self-extending AI agent can create or install new bounded capabilities, modes, or integrations in response to a task.
A self-replicating agent is a software agent capable of creating or initiating additional instances of itself.
Semantic code grouping organizes changed files by their functional meaning or architectural role rather than only by path or edit order.
Semantic HTML uses elements that describe the meaning and role of content, giving browsers and assistive technologies useful structural information.
A semantic layer gives AI systems and people a shared, governed meaning for business data such as metrics, dimensions, and relationships.