Bandits with Preference Feedback: A Stackelberg Game Perspective

arXiv:2406.16745v1 Announce Type: new
Abstract: Bandits with preference feedback present a powerful tool for optimizing unknown target functions when only pairwise comparisons are allowed instead of direct value queries. This model allows for incorporating human feedback into online inference and optimization and has been employed in systems for fine-tuning large language models. The problem is well understood in simplified settings with linear target functions or over finite small domains that limit practical interest. Taking the next step, we consider infinite …

abstract arxiv cs.ai cs.gt cs.lg feedback fine-tuning fine-tuning large language models functions game human human feedback inference language language models large language large language models optimization perspective queries stat.ml systems tool tuning type value