Unverified50% confidenceDecision RuleExact time
NoteBank's writing stage takes place during context encoding, before any question appears, selecting notes based on each token's key vector uniqueness (1 minus the cosine similarity between the token's key and the block's average key, then averaged over layers and heads), and cannot foresee what future questions will need.
1
Sources
50%
Confidence
Long-term
Relevance
10/15/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn NoteBank's replay stage, the retrieved notes are concatenated in original document order and reprocessed starting from the unmodified backbone network (starting from a saved copy of the state), after which the question is processed and the answer generated.68% similarUnverifiedWhether prompt caching hits depends on byte-for-byte consistency of the request prefix; if even one character at the beginning differs, the entire cache is invalidated.66% similarUnverifiedChanges to the tokenizer's vocabulary size, special token definitions, or context handling methods will cause some preprocessing and postprocessing logic to need updating.65% similarUnverifiedLanguage models process tokens rather than numbers, so they can accurately copy out amounts yet are prone to errors in operations such as summation, averaging and percentages, and the result for the same question may differ from one run to the next.63% similarUnverified自回归解码具有串行特性，每个Token的生成依赖前一个Token，难以并行63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1040233API
curl https://kongchang.com/api/v1/knowledge/claims/1040233MCP
get_claim(id=1040233)