Selective Regenerative Decoding: Trajectory-Level Intervention for Inference-Time Reasoning

arXiv:2608.24338 · cs.AI · Submitted 2026-08-25 · Read on arXiv

cs.AI

Submitted: 2026-08-25

Updated: 2026-08-25

Comments: Large Language Model, Test-Time Decoding Technique, Reasoning, 20 pages; Submitted to ARR

Code: https://github.com/tatsu-lab/alpaca_eval

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

Terminology

• Accuracy — 38× in this paper · explained in 1 episode
• Sample Efficiency — 14× in this paper · explained in 4 episodes
• Selective Regenerative Decoding — 10× in this paper
• Benchmark — 8× in this paper · explained in 2 episodes
• Speculative Decoding — 8× in this paper · explained in 2 episodes
• Ablation — 5× in this paper · explained in 1 episode
• Generative Model — 5× in this paper · explained in 2 episodes
• Reward-guided Speculative Decoding — 4× in this paper
• Ablation Studies — 2× in this paper · explained in 1 episode
• Alignment — 2× in this paper · explained in 1 episode
• Calibration — 2× in this paper · explained in 7 episodes
• Controlled Decoding — 2× in this paper
• RAG — 2× in this paper · explained in 3 episodes
• Transformer — 2× in this paper · explained in 2 episodes
• Action Space — 1× in this paper · explained in 1 episode
• Pruning — 1× in this paper · explained in 4 episodes
• Reward Function — 1× in this paper · explained in 2 episodes
• Sample Complexity — 1× in this paper · explained in 15 episodes

Sources

• It's MBR All the Way Down: Modern Generation Techniques Through the Lens of Minimum Bayes Risk
• WebGPT: Browser-assisted question-answering with human feedback
• The Curious Case of Neural Text Degeneration
• Learning to summarize from human feedback
• Faithfulness-Aware Decoding Strategies for Abstractive Summarization
• Measuring and Mitigating Constraint Violations of In-Context Learning for Utterance-to-API Semantic Parsing
• Efficient Guided Generation for Large Language Models

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
• 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
• VSAL: A Vision Solver with Adaptive Layouts for Graph Property Detection