Verified65% confidenceFactTime unknown
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
3
Sources
65%
Confidence
Long-term
Relevance
5/31/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedEarly large language models suffered from serious overconfidence, tending to give definitive statements even for uncertain answers, which is the hallucination phenomenon.81% similarUnverified大语言模型在涉及高度专业的化学、生物或网络安全操作时，极易产生听起来可信但存在关键错误的幻觉输出79% similarUnverifiedMainstream large language models show significantly reduced accuracy when understanding obfuscated bundles with extremely long contexts (tens of thousands of lines) and may hallucinate on custom encryption variants.76% similarVerifiedLarge language models are prone to forgetting early information or hallucinating during long conversations, which in programming manifests as fabricating non-existent APIs or generating logically incorrect code.76% similarUnverifiedLarge language models can simultaneously maintain dozens or even hundreds of independent fake identities in parallel while preserving consistency across long-term interactions, enabling large-scale persona impersonation.76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/12953API
curl https://kongchang.com/api/v1/knowledge/claims/12953MCP
get_claim(id=12953)