Unverified50% confidenceOpinionExact time
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
1
Sources
50%
Confidence
Long-term
Relevance
7/14/2026
First Seen
Sources
AI欺骗行为：为何无需意识同样危险
Related Claims
VerifiedResearch indicates that hallucination rates do not have a linear negative correlation with model scale, and larger models sometimes produce harder-to-detect misinformation in a more fluent manner.83% similarUnverifiedResearch shows that hallucination rate is negatively correlated with the knowledge density of the prompt; the richer and more precise the context provided in the prompt, the lower the probability that the model will generate fabricated content.79% similarUnverifiedPure large language models have two major deficiencies: a knowledge cutoff problem and a hallucination problem.77% similarUnverifiedLarge models output the most likely next word rather than precise retrieval of facts, which leads to hallucinations in domains with sparse coverage in the training data.77% similarUnverifiedOverly verbose, contradictory, or ambiguous system prompts cause the model's attention to become scattered, resulting in over-thinking or instruction-following failures; streamlining agents.md can improve the determinism of instruction following.76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/504375API
curl https://kongchang.com/api/v1/knowledge/claims/504375MCP
get_claim(id=504375)