Unverified60% confidenceOpinionExact time
A piece of code is read far more times than it is written.
2
Sources
60%
Confidence
Long-term
Relevance
9/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn large model invocations, text and code in both input and output are converted into token counts.69% similarUnverifiedPrompt token reading can be processed in batches in parallel, but output generation can only proceed token by token, so the read rate and the write rate are two different numbers.67% similarUnverifiedIn long-text generation scenarios, the more tokens that are generated, the greater the cumulative speedup benefit of speculative decoding.67% similarUnverifiedIn this project, the number of times context was re-read (cache reads) was approximately 14 times the number of initial writes (inputs).66% similarUnverifiedThe cached input volume is far higher than the output volume because each self-check round requires feeding a large amount of existing context back into the model.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/911230API
curl https://kongchang.com/api/v1/knowledge/claims/911230MCP
get_claim(id=911230)