Unverified50% confidenceConceptExact time
The embedding model only sees the sentence pattern of the paragraph and cannot perceive its heading; this is not a flaw of the embedding model, but rather that any fixed-length vector compression discards some information—the problem is that structural clues already provided by the author are being actively discarded
1
Sources
50%
Confidence
Long-term
Relevance
10/13/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedFor very long documents, a model may have processed only part of the content yet provide an answer that appears complete without any indication.75% similarUnverifiedUnigram 模型由 SentencePiece 库实现，从大词表出发逐步剔除对整体似然贡献最小的子词来缩减词表75% similarUnverifiedAbstractive summarization may introduce hallucinations, where the model fabricates content that does not exist in the original text.73% similarUnverifiedExtractive summarization is constrained to sentences from the source text and cannot perform linguistic reorganization and compression the way abstractive summarization can.72% similarUnverifiedPrompt injection exploits not a vulnerability in the code parser, but the model's own insufficient awareness of the boundary between instructions and data71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1019578API
curl https://kongchang.com/api/v1/knowledge/claims/1019578MCP
get_claim(id=1019578)