Unverified50% confidenceOpinionExact time
A mastery threshold is not a magic number that can be reused across models.
1
Sources
50%
Confidence
Long-term
Relevance
10/4/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedA single throughput figure (tok/s) cannot reflect a model's true capability, long-context cost, or cross-model comparability.68% similarUnverified哈达玛变换用于量化预处理时不会引入额外的数值不稳定性，因为其条件数恒为167% similarUnverifiedWhen model scale exceeds a certain threshold, certain capabilities suddenly jump from near zero to a significant level, rather than growing linearly with parameter count.65% similarUnverifiedHow to define reasonable cohort-level thresholds on small-sample, high-weight slices remains an open and unsolved problem65% similarUnverified浮点数在token空间中并非连续分布，导致LLM精确运算天然不可靠64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/971859API
curl https://kongchang.com/api/v1/knowledge/claims/971859MCP
get_claim(id=971859)