Imagen de Portada

PSD: Pushing the Pareto Frontier of Diffusion LLMs via Parallel Speculative Decoding

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Sun, Shengyin, Li, Yiming, Liu, Renxi, Li, Xinqi, Zhen, Hui-Ling, Lin, Weizhe, Chen, Chen, Yu, Xianzhi, Yuan, Mingxuan, Ma, Chen
Formato:	Preprint
Publicado:	2026
Materias:	Computation and Language
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Revisiting Judge Decoding from First Principles via Training-Free Distributional Divergence
por: Sun, Shengyin, et al.
Publicado: (2026)

Scaling Up, Speeding Up: A Benchmark of Speculative Decoding for Efficient LLM Test-Time Scaling
por: Sun, Shengyin, et al.
Publicado: (2025)

Towards Efficient Agents: A Co-Design of Inference Architecture and System
por: Lin, Weizhe, et al.
Publicado: (2025)

SwiftMem: Fast Agentic Memory via Query-aware Indexing
por: Tian, Anxin, et al.
Publicado: (2026)

Hyperbolic Contrastive Learning with Model-augmentation for Knowledge-aware Recommendation
por: Sun, Shengyin, et al.
Publicado: (2025)

AdaTSQ: Pushing the Pareto Frontier of Diffusion Transformers via Temporal-Sensitivity Quantization
por: Zhang, Shaoqiu, et al.
Publicado: (2026)

TrimR: Verifier-based Training-Free Thinking Compression for Efficient Test-Time Scaling
por: Lin, Weizhe, et al.
Publicado: (2025)

Enhanced Pre-training of Graph Neural Networks for Million-Scale Heterogeneous Graphs
por: Sun, Shengyin, et al.
Publicado: (2025)

DLLM Agent: See Farther, Run Faster
por: Zhen, Huiling, et al.
Publicado: (2026)

What Matters For Safety Alignment?
por: Li, Xing, et al.
Publicado: (2026)

SVDq: 1.25-bit and 410x Key Cache Compression for LLM Attention
por: Yankun, Hong, et al.
Publicado: (2025)

Pushing Forward Pareto Frontiers of Proactive Agents with Behavioral Agentic Optimization
por: Yao, Yihang, et al.
Publicado: (2026)

IndustryEQA: Pushing the Frontiers of Embodied Question Answering in Industrial Scenarios
por: Li, Yifan, et al.
Publicado: (2025)

EasySpec: Layer-Parallel Speculative Decoding for Efficient Multi-GPU Utilization
por: Wu, Yize, et al.
Publicado: (2025)

PEARL: Parallel Speculative Decoding with Adaptive Draft Length
por: Liu, Tianyu, et al.
Publicado: (2024)

OmniPSD: Layered PSD Generation with Diffusion Transformer
por: Liu, Cheng, et al.
Publicado: (2025)

dParallel: Learnable Parallel Decoding for dLLMs
por: Chen, Zigeng, et al.
Publicado: (2025)

DFlash: Block Diffusion for Flash Speculative Decoding
por: Chen, Jian, et al.
Publicado: (2026)

Large Language Models as Topological Structure Enhancers for Text-Attributed Graphs
por: Sun, Shengyin, et al.
Publicado: (2023)

AgentCollab: A Self-Evaluation-Driven Collaboration Paradigm for Efficient LLM Agents
por: Gao, Wenbo, et al.
Publicado: (2026)

Nearest Neighbor Speculative Decoding for LLM Generation and Attribution
por: Li, Minghan, et al.
Publicado: (2024)

DMax: Aggressive Parallel Decoding for dLLMs
por: Chen, Zigeng, et al.
Publicado: (2026)

Speculative Pipeline Decoding: Higher-Accruacy and Zero-Bubble Speculation via Pipeline Parallelism
por: Yu, Yijiong, et al.
Publicado: (2026)

SpecExec: Massively Parallel Speculative Decoding for Interactive LLM Inference on Consumer Devices
por: Svirschevski, Ruslan, et al.
Publicado: (2024)

FocuSFT: Bilevel Optimization for Dilution-Aware Long-Context Fine-Tuning
por: Pei, Zehua, et al.
Publicado: (2026)

From Pruning to Grafting: Dynamic Knowledge Redistribution via Learnable Layer Fusion
por: Pei, Zehua, et al.
Publicado: (2024)

ParallelSpec: Parallel Drafter for Efficient Speculative Decoding
por: Xiao, Zilin, 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)

TS-DP: Reinforcement Speculative Decoding For Temporal Adaptive Diffusion Policy Acceleration
por: Li, Ye, et al.
Publicado: (2025)

SpecVLM: Enhancing Speculative Decoding of Video LLMs via Verifier-Guided Token Pruning
por: Ji, Yicheng, et al.
Publicado: (2025)

MemDLM: Memory-Enhanced DLM Training
por: Pei, Zehua, et al.
Publicado: (2026)

DeepScientist: Advancing Frontier-Pushing Scientific Findings Progressively
por: Weng, Yixuan, et al.
Publicado: (2025)

Behavioral Fingerprinting of Large Language Models
por: Pei, Zehua, et al.
Publicado: (2025)

Decoding at the Speed of Thought: Harnessing Parallel Decoding of Lexical Units for LLMs
por: Sun, Chenxi, et al.
Publicado: (2024)

Pushing the Frontiers of Self-Distillation Prototypes Network with Dimension Regularization and Score Normalization
por: Chen, Yafeng, et al.
Publicado: (2025)

DART: Diffusion-Inspired Speculative Decoding for Fast LLM Inference
por: Liu, Fuliang, et al.
Publicado: (2026)

Unleashing Low-Bit Inference on Ascend NPUs: A Comprehensive Evaluation of HiFloat Formats
por: Zhao, Pengxiang, et al.
Publicado: (2026)

DiffuSpec: Unlocking Diffusion Language Models for Speculative Decoding
por: Li, Guanghao, et al.
Publicado: (2025)

Cacheback: Speculative Decoding With Nothing But Cache
por: Ma, Zhiyao, et al.
Publicado: (2025)

PASER: Post-Training Data Selection for Efficient Pruned Large Language Model Recovery
por: He, Bowei, et al.
Publicado: (2025)