Research
LLM Prompt Duel Optimizer: Efficient Label-Free Prompt Optimization
arXiv:2510.13907v3 Announce Type: replace Abstract: Large language models (LLMs) are highly sensitive to prompts, but most automatic prompt optimization (APO) methods assume access to ground-truth ref
arXiv:2510.13907v3 Announce Type: replace Abstract: Large language models (LLMs) are highly sensitive to prompts, but most automatic prompt optimization (APO) methods assume access to ground-truth references (e.g., labeled validation data) that are costly to obtain. We propose the Prompt Duel Optimizer (PDO), a sample-efficient framework for label-free prompt optimization based on pairwise preference feedback from an LLM judge. PDO casts prompt selection as a dueling-bandit problem and combines (i) Double Thompson Sampling to prioritize informative comparisons under a fixed judge budget, with (ii) top-performer guided mutation to expand the candidate pool while pruning weak prompts. Experiments on BIG-bench Hard (BBH) and MS MARCO show that PDO consistently identifies stronger prompts than label-free baselines, while offering favorable quality--cost trade-offs under constrained comparison budgets.
Related
- Adaptive Prompt Structure Factorization: A Framework for Self-Discovering and Optimizing Compositional Prompt Programs
- Towards Hierarchical Multi-Step Reward Models for Enhanced Reasoning in Large Language Models
- Rectifying LLM Thought from Lens of Optimization
- SeLaR: Selective Latent Reasoning in Large Language Models
Source: arXiv cs.CL | 2026-04-10