Imagen de Portada

Draft, Verify, and Improve: Toward Training-Aware Speculative Decoding

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Bhansali, Shrenik, Heck, Larry
Formato:	Preprint
Publicado:	2025
Materias:	Machine Learning
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

LEGO: Language Model Building Blocks
por: Bhansali, Shrenik, et al.
Publicado: (2024)

BudgetDraft: Acceptance-Aware Multi-View Training for Sparse-KV Speculative Decoding
por: He, Liang, et al.
Publicado: (2026)

DREAM-S: Speculative Decoding with Searchable Drafting and Target-Aware Refinement for Multimodal Generation
por: Liu, Zining, et al.
Publicado: (2026)

FastEagle: Cascaded Drafting for Accelerating Speculative Decoding
por: Huang, Haiduo, et al.
Publicado: (2025)

Training Domain Draft Models for Speculative Decoding: Best Practices and Insights
por: Hong, Fenglu, et al.
Publicado: (2025)

When Drafts Evolve: Speculative Decoding Meets Online Learning
por: Qian, Yu-Yang, et al.
Publicado: (2026)

Draft Less, Retrieve More: Hybrid Tree Construction for Speculative Decoding
por: Shen, Yuhao, et al.
Publicado: (2026)

POSS: Position Specialist Generates Better Draft for Speculative Decoding
por: Huang, Langlin, et al.
Publicado: (2025)

Beat the long tail: Distribution-Aware Speculative Decoding for RL Training
por: Shao, Zelei, et al.
Publicado: (2025)

TABED: Test-Time Adaptive Ensemble Drafting for Robust Speculative Decoding in LVLMs
por: Lee, Minjae, et al.
Publicado: (2026)

SlimSpec: Low-Rank Draft LM-Head for Accelerated Speculative Decoding
por: Plaksin, Anton, et al.
Publicado: (2026)

ML-SpecQD: Multi-Level Speculative Decoding with Quantized Drafts
por: Georganas, Evangelos, et al.
Publicado: (2025)

Speculative Decoding with CTC-based Draft Model for LLM Inference Acceleration
por: Wen, Zhuofan, et al.
Publicado: (2024)

Balancing Coverage and Draft Latency in Vocabulary Trimming for Faster Speculative Decoding
por: Shoham, Ofir Ben
Publicado: (2026)

Speculative Safety-Aware Decoding
por: Wang, Xuekang, et al.
Publicado: (2025)

OmniDraft: A Cross-vocabulary, Online Adaptive Drafter for On-device Speculative Decoding
por: Ramakrishnan, Ramchalam Kinattinkara, et al.
Publicado: (2025)

MineDraft: A Framework for Batch Parallel Speculative Decoding
por: Tang, Zhenwei, et al.
Publicado: (2026)

LongSpec: Long-Context Lossless Speculative Decoding with Efficient Drafting and Verification
por: Yang, Penghui, et al.
Publicado: (2025)

Direct Alignment of Draft Model for Speculative Decoding with Chat-Fine-Tuned LLMs
por: Goel, Raghavv, et al.
Publicado: (2024)

Speculative Speculative Decoding
por: Kumar, Tanishq, et al.
Publicado: (2026)

D-PACE: Dynamic Position-Aware Cross-Entropy for Parallel Speculative Drafting
por: Wu, Tianyu, et al.
Publicado: (2026)

Make Every Draft Count: Hidden State based Speculative Decoding
por: Chen, Yuetao, et al.
Publicado: (2026)

Exploring and Improving Drafts in Blockwise Parallel Decoding
por: Kim, Taehyeon, et al.
Publicado: (2024)

Cascade Speculative Drafting for Even Faster LLM Inference
por: Chen, Ziyi, et al.
Publicado: (2023)

Decoding Speculative Decoding
por: Yan, Minghao, et al.
Publicado: (2024)

Fail Fast, Win Big: Rethinking the Drafting Strategy in Speculative Decoding via Diffusion LLMs
por: Pan, Rui, et al.
Publicado: (2025)

DEER: Draft with Diffusion, Verify with Autoregressive Models
por: Cheng, Zicong, et al.
Publicado: (2025)

MARS: Unleashing the Power of Speculative Decoding via Margin-Aware Verification
por: Song, Jingwei, et al.
Publicado: (2026)

Dynamic Delayed Tree Expansion For Improved Multi-Path Speculative Decoding
por: Thomas, Rahul, et al.
Publicado: (2026)

Global Resolution: Optimal Multi-Draft Speculative Sampling via Convex Minimization
por: Thomas, Rahul Krishna, et al.
Publicado: (2025)

Weight-of-Thought Reasoning: Exploring Neural Network Weights for Enhanced LLM Reasoning
por: Punjwani, Saif, et al.
Publicado: (2025)

Co2PO: Coordinated Constrained Policy Optimization for Multi-Agent RL
por: Patel, Shrenik, et al.
Publicado: (2026)

Reject Only Critical Tokens: Pivot-Aware Speculative Decoding
por: Ziashahabi, Amir, et al.
Publicado: (2025)

PRISM: Parametrically Refactoring Inference for Speculative Sampling Draft Models
por: Wang, Xuliang, et al.
Publicado: (2026)

Hydra: Sequentially-Dependent Draft Heads for Medusa Decoding
por: Ankner, Zachary, et al.
Publicado: (2024)

FastGRPO: Accelerating Policy Optimization via Concurrency-aware Speculative Decoding and Online Draft Learning
por: Zhang, Yizhou, et al.
Publicado: (2025)

AdaEDL: Early Draft Stopping for Speculative Decoding of Large Language Models via an Entropy-based Lower Bound on Token Acceptance Probability
por: Agrawal, Sudhanshu, et al.
Publicado: (2024)

Free Draft-and-Verification: Toward Lossless Parallel Decoding for Diffusion Large Language Models
por: Wu, Shutong, et al.
Publicado: (2025)

Pruner: A Draft-then-Verify Exploration Mechanism to Accelerate Tensor Program Tuning
por: Qiao, Liang, et al.
Publicado: (2024)

Speculative Decoding Across Languages
por: Paudel, Nirajan, et al.
Publicado: (2026)