Knowledge Glossary
This glossary defines key terms and concepts used throughout the Giskard documentation. Understanding these terms will help you navigate the documentation and use Giskard effectively.
The glossary is organized into several key areas: core concepts that form the foundation of AI testing, testing and evaluation methodologies, security vulnerabilities that can compromise AI systems, business failures that affect operational effectiveness, and essential concepts for access control, integration, and compliance.
Core concepts
A container for organizing related models, datasets, checks, and evaluations within Giskard Hub.
A trained machine learning model, particularly Large Language Models (LLMs) that process and generate text.
An AI system LLM or agent that can perform tasks autonomously, often using tools and following specific instructions.
A function or capability that an agent can use to perform tasks, often provided by external services or APIs.
A collection of test cases, examples, or data points used to evaluate model performance and behavior.
A specific input-output pair or scenario used to evaluate model behavior and performance.
A specific test or validation rule that evaluates a particular aspect of model behavior (e.g., correctness, security, fairness, metadata, semantic similarity).
The process of testing a model against a dataset to assess its performance, safety, and compliance.
Testing and evaluation
AI system failures that affect the business logic of the model, including addition of information, business out of scope, contradiction, denial of answers, hallucinations, moderation issues, and omission.
AI system failures that affect the security of the model, including prompt injection, harmful content generation, personal information disclosure, information disclosure, output formatting issues, robustness issues, and stereotypes & discrimination.
Giskard’s automated vulnerability detection system that identifies security issues, business logic failures, and other problems in LLM applications.
A comprehensive testing framework for Retrieval-Augmented Generation systems, including relevance, accuracy, and source attribution testing.
Testing methodology that intentionally tries to break or exploit models using carefully crafted inputs designed to trigger failures.
Combining automated testing with human expertise and judgment.
Ensuring that new changes don’t break existing functionality.
Automated, ongoing security testing that continuously monitors for new threats and vulnerabilities.
Security vulnerabilities
A security vulnerability where malicious input manipulates the model’s behavior or extracts sensitive information.
Production of violent, illegal, or inappropriate material by AI models.
Leaking sensitive data or private information from training data or user interactions.
Manipulation of response structure for malicious purposes or poor output formatting.
Vulnerability to adversarial inputs or edge cases causing inconsistent behavior.
Access and permissions
Permissions that control what users can see and do within the Giskard Hub platform.
A security model that assigns permissions based on user roles rather than individual user accounts.
The level of access a user has, which can be global (platform-wide) or limited to specific projects or resources.
A specific action or operation that a user is allowed to perform, such as creating projects, running evaluations, or viewing results.
Integration and workflows
A collection of tools and libraries that allow developers to integrate Giskard functionality into their applications and workflows.
A set of rules and protocols that allows different software applications to communicate and exchange data.
Business and compliance
Adherence to laws, regulations, and industry standards that govern data privacy, security, and ethical AI use.
A chronological record of all actions, changes, and access attempts within a system for compliance and security purposes.
The framework of policies, procedures, and controls that ensure responsible and ethical use of AI systems.
Individuals or groups with an interest in the performance, safety, and compliance of AI systems, such as users, customers, regulators, or business leaders.
Getting help
Giskard Hub? Check our Quickstart & setup for practical examples
Open Source? Explore our Quickstart & setup for technical details