Expired50% confidenceFactTime unknown
Top-k sampling restricts the word selection range to the k words with the highest probabilities in order to prevent absurd outputs
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 9/30/2026(expired)
Sources
Demystifying Transformer: A Word-Continuation Function, Deconstructed
bilibiliAI算法工程师Fiona
Related Claims
UnverifiedTop-p (nucleus sampling) limits the candidate token range by dynamically truncating probability mass, and has better adaptability than fixed-truncation Top-k75% similarUnverifiedTop-K的K值存在经典权衡：K太小会漏掉关键信息，K太大会引入噪声导致大模型答案跑偏74% similarUnverifiedThe choice of K value in Top-K routing affects the trade-off between computational efficiency and expressive capacity — a smaller K leads to faster inference but may sacrifice expressiveness, and in practice K is typically set to 1 or 2.69% similarUnverifiedCoze's Top-P nucleus sampling mechanism starts from the highest-probability words and accumulates probabilities until the cumulative probability reaches the set P value, sampling only from those words.67% similarUnverifiedRAG调优中Top K（返回最相关的K个片段）和相似度阈值是最常需要调整的两个参数67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/54656API
curl https://kongchang.com/api/v1/knowledge/claims/54656MCP
get_claim(id=54656)