Unverified50% confidenceFactExact time
Perplexity能够提供带引用来源的答案，部分解决了大语言模型的幻觉问题
1
Sources
50%
Confidence
Long-term
Relevance
8/21/2026
First Seen
Sources
Related Claims
UnverifiedLarge language models have a miscalibration problem, often answering in a confident tone regardless of whether the answer is correct or incorrect.79% similarUnverified困惑度（Perplexity）是自然语言处理领域衡量语言模型对文本意外程度的核心指标，困惑度越低文本越可预测75% similarUnverifiedText generated by large language models tends to have low perplexity, with word choices that are overly consistent with probabilistic expectations, lacking the unexpectedness of human writing.74% similarUnverifiedEarly large language models suffered from serious overconfidence, tending to give definitive statements even for uncertain answers, which is the hallucination phenomenon.72% similarUnverified当需求存在逻辑矛盾或业务知识缺失时，大语言模型因缺乏真实世界反馈机制存在产生幻觉（Hallucination）的风险72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/782860API
curl https://kongchang.com/api/v1/knowledge/claims/782860MCP
get_claim(id=782860)