Unverified60% confidenceFactExact time
Expressions like "tired" in model output occur because training data is filled with human modes of expression, representing pattern matching at the linguistic level rather than genuine state feedback.
2
Sources
60%
Confidence
Long-term
Relevance
9/15/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedWhen a model's training data contains benchmark test questions or highly similar content, the model may achieve high scores through memorization rather than genuine understanding.80% similarVerifiedWhen input information is incomplete, large language models complete it based on the most common patterns in their training data, which is statistically called maximum likelihood estimation.78% similarUnverified模型对 schema 的遵循能力受训练数据分布影响，训练语料中高频出现的字段组合会导致模型产生幻觉字段77% similarUnverifiedWhen a language model is asked to show its thinking, what it outputs is a self-description reorganized from linguistic patterns related to relevant concepts in its training data, not a genuine readout of its internal states.76% similarVerifiedHallucinations stem from the model's optimization objective itself—the model is trained to generate linguistically coherent and statistically plausible continuations, not factually accurate answers.76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/923823API
curl https://kongchang.com/api/v1/knowledge/claims/923823MCP
get_claim(id=923823)