LibraSpec: Dynamic Diffusion-Based Speculative Decoding via Marginal-Gain-Driven Optimization

arXiv:2608.08721 · cs.CL, cs.AI · Submitted 2026-08-09 · Read on arXiv

Zexun Lin, Yuan Feng, Junlin Lv, Kevin S. Zhou, Xike Xie

cs.CL, cs.AI

Submitted: 2026-08-09

Updated: 2026-08-11

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Speculative Decoding — 53× in this paper · explained in 2 episodes
• Verification — 30× in this paper · explained in 2 episodes
• Benchmark — 20× in this paper · explained in 2 episodes
• Calibration — 4× in this paper · explained in 9 episodes
• Adam — 3× in this paper · explained in 3 episodes
• Block Diffusion — 2× in this paper · explained in 2 episodes
• Diffusion Model — 2× in this paper · explained in 10 episodes
• Iterative Refinement — 2× in this paper · explained in 5 episodes
• Thinking mode — 2× in this paper · explained in 2 episodes
• Transformer — 2× in this paper · explained in 2 episodes
• Transformers — 2× in this paper · explained in 2 episodes
• Computational overhead — 1× in this paper · explained in 2 episodes
• Confidence Calibration — 1× in this paper · explained in 2 episodes
• Diffusion Models — 1× in this paper · explained in 9 episodes
• Generalization — 1× in this paper · explained in 14 episodes
• Hyperparameters — 1× in this paper · explained in 2 episodes
• Information Gain — 1× in this paper · explained in 2 episodes
• LLM-as-a-Judge — 1× in this paper · explained in 4 episodes
• Language Models (LLMs) — 1× in this paper · explained in 2 episodes
• Large Language Models (LLMs) — 1× in this paper · explained in 9 episodes
• Papers — 1× in this paper · explained in 3 episodes
• Performance Gap — 1× in this paper · explained in 2 episodes
• Reinforcement Learning — 1× in this paper · explained in 4 episodes
• Robustness — 1× in this paper · explained in 6 episodes
• Sufficiency — 1× in this paper · explained in 2 episodes
• Verifier — 1× in this paper · explained in 2 episodes

Sources

• Medusa: Simple LLM Inference Acceleration Framework with Multiple Decoding Heads
• DFlash: Block Diffusion for Flash Speculative Decoding
• Evaluating Large Language Models Trained on Code
• Speculative Diffusion Decoding: Accelerating Language Generation through Diffusion
• Training Verifiers to Solve Math Word Problems
• BanditSpec: Adaptive Speculative Decoding via Bandit Algorithms
• JetSpec: Breaking the Scaling Ceiling of Speculative Decoding with Parallel Tree Drafting
• SpecDec++: Boosting Speculative Decoding via Adaptive Candidate Lengths
• Speculative Verification: Exploiting Information Gain to Refine Speculative Decoding
• Efficient Memory Management for Large Language Model Serving with PagedAttention
• Fast Inference from Transformers via Speculative Decoding
• DiffuSpec: Unlocking Diffusion Language Models for Speculative Decoding
• SpecForge: A Flexible and Efficient Open-Source Training Framework for Speculative Decoding
• EAGLE-2: Faster Inference of Language Models with Dynamic Draft Trees
• EAGLE-3: Scaling up Inference Acceleration of Large Language Models via Training-Time Test
• EAGLE: Speculative Sampling Requires Rethinking Feature Uncertainty
• Let's Verify Step by Step
• DART: Diffusion-Inspired Speculative Decoding for Fast LLM Inference
• Dynamic Speculation Lookahead Accelerates Speculative Decoding of Large Language Models
• SSSD: Simply-Scalable Speculative Decoding

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering