Skip to content

Repository files navigation

Speculative Decoding Research Knowledge Base banner

Speculative Decoding 研究知识库

把论文森林变成一张可比较、可复现、可以直接选题的研究地图。

Documentation Audit DSpark Latest scan

66 篇核心精读 · 1,188 页全文核读 · 6 条研究主线 · 最新检索至 2026-08-26


Note

66 篇核心精读快照截止 2026-08-10;2026-08-26 增量更新另核验 20 份 arXiv PDF、补出 4 个正式来源方法。增量项尚未冒充逐页精读完成。每个核心条目均记录已读版本、页码范围与 PDF SHA-256;原始 PDF 因版权不进入仓库。

这个知识库面向准备进入 speculative decoding 研究的读者。目标不是复述摘要,而是把每篇论文的问题、假设、算法、公式、训练与推理流程、正确性边界、实验、实现路径、复现风险、局限和可继续研究的问题压缩进一个可独立阅读的中文文件。

打开可搜索文档站 →

从哪里开始

阅读状态与证据边界

本 README 只列出 deep_read_complete 条目。原始 PDF 与抽取文本用于本地核读,因版权不进入仓库;公开文件保留官方入口、版本、页码和哈希。速度数字均按原论文硬件、batch、temperature、backend和baseline解释,不把最高 endpoint 当作普遍结论。详见 调研与精读方法来源清单

核心论文目录

01 基础、理论与综述(9 篇)

查看本类导读与推荐阅读路线

年份 论文 Venue
2018 Blockwise Parallel Decoding for Deep Autoregressive Models NeurIPS 2018
2022 Speculative Decoding: Exploiting Speculative Execution for Accelerating Seq2seq Generation arXiv / ICLR 2023 submission
2023 Accelerating Large Language Model Decoding with Speculative Sampling arXiv technical report
2023 Fast Inference from Transformers via Speculative Decoding ICML 2023
2023 SpecTr: Fast Speculative Decoding via Optimal Transport NeurIPS 2023
2024 Unlocking Efficiency in Large Language Model Inference: A Comprehensive Survey of Speculative Decoding Findings of ACL 2024
2025 Decoding Speculative Decoding NAACL 2025
2026 Global Resolution: Optimal Multi-Draft Speculative Sampling via Convex Minimization ICLR 2026 Oral
2026 When Is a Draft Accepted? A Theory of Acceptance in Speculative Decoding arXiv preprint

02 独立 drafter、对齐与在线选择(7 篇)

查看本类导读与推荐阅读路线

年份 论文 Venue
2023 Accelerating LLM Inference with Staged Speculative Decoding ICML 2023 workshop / arXiv
2023 Speculative Decoding with Big Little Decoder NeurIPS 2023
2024 DistillSpec: Improving Speculative Decoding via Knowledge Distillation ICLR 2024
2024 Online Speculative Decoding ICML 2024
2025 Learning Harmonized Representations for Speculative Sampling ICLR 2025
2026 Not-a-Bandit: Provably No-Regret Drafter Selection in Speculative Decoding for LLMs ICLR 2026
2026 Speculative Decoding and the Curse of Multilinguality arXiv preprint

03 Feature head、MTP 与并行块草稿(20 篇)

查看本类导读与推荐阅读路线

年份 论文 Venue
2024 EAGLE-2: Faster Inference of Language Models with Dynamic Draft Trees EMNLP 2024
2024 EAGLE: Speculative Sampling Requires Rethinking Feature Uncertainty ICML 2024
2024 Hydra: Sequentially-Dependent Draft Heads for Medusa Decoding COLM 2024
2024 MEDUSA: Simple LLM Inference Acceleration Framework with Multiple Decoding Heads ICML 2024
2024 Recurrent Drafter for Fast Speculative Decoding in Large Language Models arXiv preprint
2025 EAGLE-3: Scaling up Inference Acceleration of Large Language Models via Training-Time Test NeurIPS 2025
2025 PARD: Accelerating LLM Inference with Low-Cost Parallel Draft Model Adaptation arXiv preprint
2026 AngelSpec: Towards Real-World High Performance Inference with Speculative Decoding arXiv preprint
2026 CURE: Local Uncertainty Repair for Block-Parallel Speculative Decoding arXiv preprint
2026 DBLast: Dependent Block Drafting for Stochastic Speculative Decoding arXiv preprint
2026 DeLS-Spec: Decoupled Long-Short Contexts for Parallel Speculative Drafting arXiv preprint
2026 DFLARE: Scaling Up Draft Capacity for Block Diffusion Speculative Decoding arXiv preprint
2026 DFlash: Block Diffusion for Flash Speculative Decoding ICML 2026
2026 Domino: Decoupling Causal Modeling from Autoregressive Drafting in Speculative Decoding arXiv preprint
2026 DSpark: Confidence-Scheduled Speculative Decoding with Semi-Autoregressive Generation arXiv preprint
2026 From Chains to Trees: Parent-Conditioned Drafting for Semi-Autoregressive Speculative Decoding arXiv preprint
2026 JetSpec: Breaking the Scaling Ceiling of Speculative Decoding with Parallel Tree Drafting arXiv preprint
2026 P-EAGLE: Parallel-Drafting EAGLE with Scalable Training arXiv preprint
2026 TreeFlash: Parallel AR-Approximation for Faster Speculative Decoding arXiv preprint
2026 xPress: Parallel Refinement for Diffusion Drafters in Speculative Decoding arXiv preprint

04 Tree、多候选与 verification(9 篇)

查看本类导读与推荐阅读路线

年份 论文 Venue
2023 SpecInfer: Accelerating Large Language Model Serving with Tree-based Speculative Inference and Verification ASPLOS 2024
2024 Multi-Candidate Speculative Decoding arXiv preprint
2024 SEQUOIA: Scalable and Robust Speculative Decoding arXiv preprint
2024 SpecExec: Massively Parallel Speculative Decoding for Interactive LLM Inference on Consumer Devices arXiv preprint
2025 Block Verification Accelerates Speculative Decoding ICLR 2025
2025 HeteroSpec: Leveraging Contextual Heterogeneity for Efficient Speculative Decoding arXiv preprint
2026 Approximate Speculative Decoding arXiv preprint
2026 MARS: Unleashing the Power of Speculative Decoding via Margin-Aware Verification arXiv preprint
2026 Revisiting Lossy Verification in Speculative Decoding: Mechanisms, Trade-offs, and Failure Modes arXiv preprint

05 Training-free、自推测与长上下文(10 篇)

查看本类导读与推荐阅读路线

年份 论文 Venue
2024 Break the Sequential Dependency of LLM Inference Using Lookahead Decoding ICML 2024
2024 Draft & Verify: Lossless Large Language Model Acceleration via Self-Speculative Decoding ACL 2024
2024 MagicDec: Breaking the Latency-Throughput Tradeoff for Long Context Generation with Speculative Decoding ICLR 2025
2024 REST: Retrieval-Based Speculative Decoding NAACL 2024
2024 SuffixDecoding: Extreme Speculative Decoding for Emerging AI Applications NeurIPS 2025
2024 TriForce: Lossless Acceleration of Long Sequence Generation with Hierarchical Speculative Decoding COLM 2024
2025 LongSpec: Long-Context Lossless Speculative Decoding with Efficient Drafting and Verification arXiv preprint
2025 SpecExtend: A Drop-in Enhancement for Speculative Decoding of Long Sequences arXiv preprint
2026 Oilbird: Training-Free Speculative Decoding with Keys the Verifier Already Computes arXiv preprint
2026 Windowed-MTP: Removing the Full-Context Draft-KV Tax at Million-Token Context arXiv preprint

06 Serving、基准、安全与应用(11 篇)

查看本类导读与推荐阅读路线

年份 论文 Venue
2023 The Synergy of Speculative Decoding and Batching in Serving Large Language Models arXiv preprint
2025 Speculative Streaming: Efficient and Scalable Speculative Decoding with Multi-Stream Attention EMNLP 2025
2026 Accelerating Large-Scale Reasoning Model Inference: Self-Speculative Decoding with Sparse Attention (SparseSpec) MLSys 2026
2026 AcceptMoE: Commitment-Weighted Self-Sizing Verifier Expert Sets for Efficient MoE Speculative Decoding arXiv preprint
2026 Adversarial Prompts for Acceptance Collapse in Speculative Decoding arXiv preprint
2026 Lossless but Not Free: An Empirical Anatomy of Speculative Decoding on Consumer Hardware arXiv preprint
2026 Mistletoe: Stealthy Acceleration-Collapse Attacks on Speculative Decoding arXiv preprint
2026 PRISM: Parametrically Refactor Inference for Speculative Decoding Draft Models MLSys 2026
2026 SpecRoll: Fast-Slow Verifier-Feedback Adaptation for Speculative Reinforcement Learning Rollouts arXiv preprint
2026 Speculative Decoding: Performance or Illusion? MLSys 2026
2026 SPEED-Bench: A Unified and Diverse Benchmark for Speculative Decoding ICML 2026

仓库结构

papers/       66 篇逐篇精读,按研究问题分为 6 类
collections/  6 个专题入口与推荐阅读路线
landscape/    全方向综述、研究空白与候选问题
metadata/     核心语料元数据与高召回候选表
assets/       文档站视觉样式、图标与横幅
scripts/      语料构建、阅读证据包与质量审计脚本

维护原则

  • 论文是否“无损”以输出序列/分布的数学保证为准,不以任务分数近似不变代替。
  • acceptance length、wall-clock speedup、throughput 与 goodput 分开记录。
  • 跨论文比较先统一硬件、engine、batch、上下文、输出长度、temperature、tree/block budget 与 baseline。
  • 新论文先进入候选表,经人工相关性筛选和全文精读后才进入核心目录。
  • 当前快照日期之后出现的论文不被暗示为已覆盖。

About

截至 2026-08-10 的投机解码研究知识库:66 篇核心论文全文精读、方法谱系、系统比较与研究空白

Topics

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages