Optimal Budgeted Adaptation of Large Language Models
This work addresses the problem of label-efficient adaptation of LLMs for practitioners, though it appears incremental as it builds on existing game-theoretic and contextual bandit methods.
The paper tackles the challenge of balancing labeled data availability and downstream accuracy in fine-tuning large language models by proposing a budget-aware supervised fine-tuning framework based on a contextual Stackelberg game, achieving regret bounds of $ ilde{O}(d\sqrt{T})$ and $ ilde{O}(\sqrt{dB} + c\sqrt{B})$ with a label-querying strategy.
The trade-off between labeled data availability and downstream accuracy remains a central challenge in fine-tuning large language models (LLMs). We propose a principled framework for \emph{budget-aware supervised fine-tuning} by casting LLM adaptation as a contextual Stackelberg game. In our formulation, the learner (leader) commits to a scoring policy and a label-querying strategy, while an adaptive environment (follower) selects challenging supervised alternatives in response. To explicitly address label efficiency, we incorporate a finite supervision budget directly into the learning objective. Our algorithm operates in the full-feedback regime and achieves $\tilde{O}(d\sqrt{T})$ regret under standard linear contextual assumptions. We extend the framework with a Largest-Latency-First (LLF) confidence gate that selectively queries labels, achieving a budget-aware regret bound of $\tilde{O}(\sqrt{dB} + c\sqrt{B})$ with $B=βT$.