---
title: "Give the Drafter a Second Chance to Guess: Parallel Reconsideration for Speculative Decoding"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2611.05724v1"
permalink: "https://feed.astralor.com/items/itm_eda5a7ef9a66603a2edcbcac"
published: "2026-11-03T22:52:34.000Z"
score: 55
channel: "tech"
tags: ["LLM","推理效率","RAG","Give","Drafter","Second Chance","Guess","Parallel Reconsideration"]
---

# Give the Drafter a Second Chance to Guess: Parallel Reconsideration for Speculative Decoding

## 速览

Speculative decoding accelerates large language model inference with a small drafter, whose per-guess acceptance rate determines the achievable speedup. Feature-level drafters share a design choice: every token, easy or…

## 关键点

- Speculative decoding accelerates large language model inference with a small drafter, whose per-guess acceptance rate determines the achievable speedup.
- Feature-level drafters share a design choice: every token, easy or hard, is drafted under a single forward pass.
- Many of the tokens the target rejects are ones the drafter is itself unsure about yet the target answers decisively.

## 全文

Speculative decoding accelerates large language model inference with a small drafter, whose per-guess acceptance rate determines the achievable speedup. Feature-level drafters share a design choice: every token, easy or hard, is drafted under a single forward pass. Many of the tokens the target rejects are ones the drafter is itself unsure about yet the target answers decisively. To address this, we introduce Parallel Reconsideration: guided by its own uncertainty, the drafter reconsiders its hesitant guesses with extra passes that run in parallel with the one-pass draft, adding negligible latency. A co-designed training recipe and serving-time drafting algorithm append these reconsidered guesses to the original draft tree as refined candidates. Across models and benchmarks, on average, Parallel Reconsideration improves over the baselines by 5.3% to 10.2% on acceptance length and by 3.5% to 9.2% on speedup.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_eda5a7ef9a66603a2edcbcac