Reinforced Efficient Reasoning via Semantically Diverse Exploration

arxiv:2601.05053

Original source · Paper JSON

Post-training · 2026-01-08

Cites (7) · Cited by (0)

Complete cleaned Markdown · Source Markdown including bibliography

Read in parts

1. Part 1
2. Part 2
3. Part 3

Contents

• Reinforced Efficient Reasoning via Semantically Diverse Exploration
• Abstract
• 1 Introduction
• 2 Related Work
• 2.1 Reinforcement Learning for LLMs
• 2.2 MCTS for LLM Reasoning
• 3 Method
• 3.1 Semantic-Entropy Guided Exploration
• 3.2 Advantage Estimation
• 3.3 Model Training
• 4 Experimental Setup
• 4.1 Datasets and Metrics
• 4.2 Model and Baselines
• 4.3 Implementation Details
• 5 Experimental Results
• 5.1 Overall Performance
• 5.2 Branching Metric Analysis
• 5.3 Reasoning Efficiency Analysis
• 5.4 Ablation Study
• 5.5 Case Study
• 5.6 Response Length Dynamics
• 5.7 Impact of ε \varepsilon -exploration
• 6 Conclusion
• Limitations
• Ethical Considerations
• Appendix A Case Study
• Instructions for reporting errors