Unverified50% confidenceFactExact time
主流LLM采用下一词预测目标，模型优化的是语言分布拟合度而非事实准确性，RLHF无法从根本上消除幻觉，完全消除幻觉在现有架构下被普遍认为不可能
1
Sources
50%
Confidence
Long-term
Relevance
7/13/2026
First Seen
Sources
Karp Speaks Bluntly: Where Does the Anger and Anxiety of CEOs in the AI Era Come From?
hackernewshackernews7/11/2026
Related Claims
VerifiedLLM产生幻觉的深层机制在于模型被优化为生成'听起来合理'的文本而非'事实准确'的文本，模型没有独立的知识检索模块78% similarUnverifiedThe core capability of modern LLMs lies in 'next token prediction,' generating coherent text through statistical language patterns rather than truly understanding semantics.75% similarUnverifiedHallucinations stem from the model's optimization objective itself—the model is trained to generate linguistically coherent and statistically plausible continuations, not factually accurate answers.75% similarUnverified主流LLM的预训练采用下一个token预测作为目标，优化的是语言建模损失而非事实准确性75% similarUnverified大模型幻觉是指LLM在生成内容时以高置信度输出看似合理但不符合用户需求的内容75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/501198API
curl https://kongchang.com/api/v1/knowledge/claims/501198MCP
get_claim(id=501198)