Unverified50% confidenceFactExact time
When a large language model is asked to "summarize" an almost empty input, it does not return "insufficient input," because its training objective is to generate the most likely continuation given the context, rather than to judge whether the information is sufficient.
1
Sources
50%
Confidence
Long-term
Relevance
10/14/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe advantage of continued pre-training is that it can exploit massive unlabeled monolingual data, but the cost is that it requires a large amount of compute, and what is learned is only the distribution at the level of language form, not directly targeted at translation, the cross-lingual mapping task.76% similarUnverifiedA language model in the pre-training stage only learns to continue text and does not understand the concept of answering questions or following instructions75% similarVerifiedLarge language models do not possess true memory; during each request, the model can only see all the text passed in with that request and is subject to a maximum context length limit.74% similarUnverifiedWhen using a model to polish or rewrite text, simply asking it to preserve the original tone and wording is often insufficient; the more fundamental approach is to rephrase the output in one's own words afterward.73% similarVerified提示注入的根本原因在于大语言模型缺乏区分指令层与数据层的原生机制73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1027708API
curl https://kongchang.com/api/v1/knowledge/claims/1027708MCP
get_claim(id=1027708)