Unverified50% confidenceFactExact time
推测性解码由Google Brain于2023年正式提出，核心洞察是GPU在验证一批Token时的计算成本与验证单个Token相近
1
Sources
50%
Confidence
Long-term
Relevance
7/9/2026
First Seen
Sources
DeepSeek DSpec Hands-On: How Speculative Decoding Accelerates LLM Inference by 6x
bilibili程序员-智能译站7/8/2026
Related Claims
UnverifiedSpeculative decoding was formally proposed by Google in 2023, using a draft model with 1/10 to 1/50 the parameters of the target model to generate multiple candidate tokens at once, which are then verified in parallel by the large model.76% similarUnverified推测解码由Google DeepMind的Leviathan等人和陈等人于2023年独立提出，基于拒绝采样的验证算法确保输出分布与主模型完全等价76% similarVerifiedThe speculative decoding paradigm was proposed almost simultaneously by teams at Google Brain and DeepMind in 2023.76% similarVerifiedSpeculative decoding was formally proposed by Google DeepMind in 2023; overall throughput can be improved by 2–3 times, and the output quality is completely equivalent to that of the original large model.73% similarUnverifiedSpeculative decoding was proposed by Google in 2022 and can achieve a 2–3x throughput improvement without sacrificing output quality72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/362544API
curl https://kongchang.com/api/v1/knowledge/claims/362544MCP
get_claim(id=362544)