entity · derived
Hallucination
Derived node: assembled mechanically from the claims carrying hallucination. A roster, not an adjudicated definition.
Every claim under this term
- 5541658-007 : Retrieval-augmented generation reduces but does not eliminate hallucination: leading legal research tools including Lexis+AI and Westlaw AI-Assisted Research still generate false citations or facts in
- 6421319-021 : If probabilistic AI agents cannot reliably distinguish truth from fabrication, the lemons problem and signaling games retain relevance, but the asymmetry relocates: it now runs between agents that can
- 6421319-022 : Hallucination rates vary sharply by domain: leading frontier models achieve sub one percent rates for general knowledge queries, while rates climb to five to thirty percent for specialized domains and
- 6421319-023 : Probabilistic language models hallucinate because they are trained to predict statistically likely token sequences rather than to verify propositional truth, so the error is intrinsic to the substrate
- 6421319-024 : Hybrid neuro symbolic architectures structurally shrink the hallucination surface because the probabilistic layer generates candidate outputs while a symbolic layer filters, validates, and rejects any