Unverified85% confidenceConceptTime unknown
大语言模型幻觉（Hallucination）在多轮对话或多步推理的Agent场景中危害被成倍放大，形成「幻觉级联」（Hallucination Cascade）
1
Sources
85%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedTraditional end-to-end LLM calls often suffer from hallucinations or logical leaps when facing multi-step, multi-constraint complex tasks.72% similarUnverifiedLarge language models are prone to hallucination drift when handling detail consistency problems that require precise long-range context tracking.71% similarUnverified模型规模越大，幻觉的表述往往越流畅、越具迷惑性，识别难度反而上升70% similarVerifiedIn multi-Agent scenarios, hallucination problems are cascadingly amplified: the erroneous output of one Agent may become the input of the next, causing errors to accumulate continuously throughout the chain.69% similarUnverifiedHallucination refers to LLMs generating plausible-sounding but factually incorrect information69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/22390API
curl https://kongchang.com/api/v1/knowledge/claims/22390MCP
get_claim(id=22390)