التقنيات

Hallucination

التعريف

The tendency of LLMs to generate plausible-sounding but factually incorrect or fabricated content. Hallucination is a primary enterprise risk for generative AI and is mitigated through RAG, grounding techniques, output validation, and human-in-the-loop review processes.

مصطلحات ذات صلة

Retrieval-Augmented Generation (RAG)

A technique that enhances LLM responses by retrieving relevant information from external knowledge sources before generating a response. RAG improves accuracy, reduces hallucinations, and enables LLMs to access up-to-date information without retraining.

Knowledge Graph

A structured representation of facts as entities and their relationships, enabling machines to reason about complex, interconnected information. Knowledge graphs complement LLMs in enterprise settings by providing verifiable, traceable facts that reduce hallucination.

Human-in-the-Loop (HITL)

A design pattern where a human must approve or review an AI system's output before it produces a real-world effect. HITL is the strongest form of human oversight and is required for the highest-stakes automated decisions under GDPR Article 22 and the EU AI Act.

خدمات ذات صلة

Eu Ai Act Compliance

Product Leadership Program

تحتاج مساعدة في فهم الذكاء الاصطناعي؟

احجز مكالمة تقييم ملاءمة Physical AI لمناقشة كيفية تطبيق مفاهيم الذكاء الاصطناعي هذه على قطاعك وتحدياتك.