mlsanityData to 7 Oct 2026

P-EAGLE: Parallel-Drafting EAGLE with Scalable Training

arXiv 2602.01469 ↗First posted 1 Feb 2026cs.LG cs.AIcited by 5

On the map: LLM reasoning → Test-time compute and efficient thinking

also in 1 more
Abstract (arXiv)

Reasoning LLMs produce longer outputs, requiring speculative decoding drafters trained on extended sequences. Parallel drafting - predicting multiple tokens per forward pass - offers latency benefits over sequential generation, but training complexity scales quadratically with the product of sequence length and parallel positions, rendering long-context training impractical. We present P(arallel)-EAGLE, which transforms EAGLE from autoregressive to parallel multi-token prediction via a learnable shared hidden state. To scale training to long contexts, we develop a framework featuring attention mask pre-computation and sequence partitioning techniques, enabling gradient accumulation within individual sequences for parallel-prediction training. We implement P-EAGLE in vLLM and demonstrate speedups of 1.10-1.36x over autoregressive EAGLE-3 across GPT-OSS 120B, 20B, and Qwen3-Coder 30B.

Key facts. Cited by 5 papers in the map (data through 7 Oct 2026). Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning. Builds on 35 of its 37 references that are in the map. First posted on arXiv 1 Feb 2026.

Builds on

35 of 37 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024EAGLE: Speculative Sampling Requires Rethinking Feature Uncertainty26 Jan

Other references · 34

  1. 2024The Llama 3 Herd of Modelscited by 14,192
  2. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  3. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,393
  4. 2021Evaluating Large Language Models Trained on Codecited by 8,053
  5. 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,767
  6. 2021RoFormer: Enhanced Transformer with Rotary Position Embeddingcited by 4,581
  7. 2020Longformer: The Long-Document Transformercited by 3,362
  8. 2023Efficient Streaming Language Models with Attention Sinkscited by 2,115
  9. 2022Fast Inference from Transformers via Speculative Decodingcited by 1,615
  10. 2022SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Modelscited by 1,338
  11. 2023SGLang: Efficient Execution of Structured Language Model Programscited by 1,107
  12. 2023Accelerating Large Language Model Decoding with Speculative Samplingcited by 975
22 more
  1. 2023Enhancing Chat Language Models by Scaling High-quality Instructional Conversationscited by 825
  2. 2024Medusa: Simple LLM Inference Acceleration Framework with Multiple Decoding Headscited by 821
  3. 2023H2O: Heavy-Hitter Oracle for Efficient Generative Inference of Large Language Modelscited by 805
  4. 2024KIVI: A Tuning-Free Asymmetric 2bit Quantization for KV Cachecited by 604
  5. 2024KVQuant: Towards 10 Million Context Length LLM Inference with KV Cache Quantizationcited by 571
  6. 2023FlexGen: High-Throughput Generative Inference of Large Language Models with a Single GPUcited by 567
  7. 2024Better & Faster Large Language Models via Multi-token Predictioncited by 344
  8. 2025EAGLE-3: Scaling up Inference Acceleration of Large Language Models via Training-Time Testcited by 318
  9. 2024Break the Sequential Dependency of LLM Inference Using Lookahead Decodingcited by 295
  10. 2023REST: Retrieval-Based Speculative Decodingcited by 148
  11. 2024TriForce: Lossless Acceleration of Long Sequence Generation with Hierarchical Speculative Decodingcited by 98
  12. 2024MagicDec: Breaking the Latency-Throughput Tradeoff for Long Context Generation with Speculative Decodingcited by 87
  13. 2023Cascade Speculative Drafting for Even Faster LLM Inferencecited by 83
  14. 2024GliDe with a CaPE: A Low-Hassle Method to Accelerate Speculative Decodingcited by 74
  15. 2024WKVQuant: Quantizing Weight and Key/Value Cache for Large Language Models Gains Morecited by 69
  16. 2024Learning Harmonized Representations for Speculative Samplingcited by 69
  17. 2025OpenCodeInstruct: A Large-scale Instruction Tuning Dataset for Code LLMscited by 57
  18. 2025PARD: Accelerating LLM Inference with Low-Cost PARallel Draft Model Adaptationcited by 28
  19. 2024ParallelSpec: Parallel Drafter for Efficient Speculative Decodingcited by 20
  20. 2024Falcon: Faster and Parallel Inference of Large Language Models through Enhanced Semi-Autoregressive Drafting and Custom-Designed Decoding Treecited by 20
  21. 2024Mixture of Attentions For Speculative Decodingcited by 17
  22. 2024BiTA: Bi-Directional Tuning for Lossless Acceleration in Large Language Modelscited by 13

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 5.

5 papers.

  1. 2026Draft-OPD: On-Policy Distillation for Speculative Draft Modelscited by 10
  2. 2026PARD-2: Target-Aligned Parallel Draft Model for Dual-Mode Speculative Decodingcited by 7
  3. 2026Online Draft Co-Training for Speculative Decoding in Large-Scale, Long-Context RL Post-Trainingcited by 2
  4. 2026Fully Rotation-Equivariant Spectral-Spatial Learning for Multispectral Object Detectioncited by 0
  5. 2026LongSpark: Efficient speculative decoding with a fixed-cost parallel draftercited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026DeLS-Spec: Decoupled Long-Short Contexts for Parallel Speculative Drafting
  2. 2025FastEagle: Cascaded Drafting for Accelerating Speculative Decoding
  3. 2026D-PACE: Dynamic Position-Aware Cross-Entropy for Parallel Speculative Drafting
  4. 2026DEdit: Iterative Draft Editing for Speculative Decoding
  5. 2026Learning to Draft: Adaptive Speculative Decoding with Reinforcement Learning
  6. 2026KnapSpec: Self-Speculative Decoding via Adaptive Layer Selection as a Knapsack Problem
  7. 2026Carryover Drafting: Recycling Rejected States for Speculative Decoding
  8. 2026Speculative Pipeline Decoding: Higher-Accuracy Drafting with Hidden Latency via Pipeline Parallelism