Where Draft Trees Lose Target Mass: Exit-Guided Speculative Decoding

arXiv:2610.11750 · cs.AI · Submitted 2026-10-08 · Read on arXiv

cs.AI

Submitted: 2026-10-08

Updated: 2026-10-08

Code: https://github.com/hsj576/TEV

Terminology

• Speculative Decoding — 1× in this paper · explained in 2 episodes

Sources

• Accelerating Large Language Model Decoding with Speculative Sampling
• Evaluating Large Language Models Trained on Code
• Recurrent Drafter for Fast Speculative Decoding in Large Language Models
• Training Verifiers to Solve Math Word Problems
• The Llama 3 Herd of Models
• Verification-Aware Training for Speculative Decoding
• Draft-OPD: On-Policy Distillation for Speculative Draft Models
• DARTree: Speculative Diffusion Decoding with Autoregressive Draft Trees
• Accelerating Speculative Decoding with Block Diffusion Draft Trees
• The Perfect Blend: Redefining RLHF with Mixture of Judges
• Qwen3 Technical Report

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• Agents' Last Exam
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems