Verified75% confidenceFactExact time
Early large language models suffered from serious overconfidence, tending to give definitive statements even for uncertain answers, which is the hallucination phenomenon.
5
Sources
75%
Confidence
Long-term
Relevance
7/16/2026
First Seen
Sources
Related Claims
Verified研究表明大语言模型规模越大，幻觉的伪装性往往越强，能生成更连贯、更具说服力的错误内容81% similarUnverifiedLanguage models generally suffer from overconfidence, especially in CoT, where erroneous steps are often written in as fluent and as confident a tone as correct steps, making them hard for readers to distinguish.80% similarVerifiedHallucination refers to large language models generating content that does not actually exist or contradicts the facts, delivered in a highly confident tone.78% similarUnverifiedLarge language models fine-tuned with RLHF tend to systematically exhibit an overconfidence bias.76% similarVerifiedCurrent mainstream large language models are generally overconfident, partly because the training objective optimizes for the fluency of predicting the next word rather than the actual reliability of the output.76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/530714API
curl https://kongchang.com/api/v1/knowledge/claims/530714MCP
get_claim(id=530714)