Black-Box APIs

ExecTune: Effective Steering of Black-Box LLMs with Guide Models

This research introduces Guide-Core Policies (GCoP), a framework for steering black-box LLMs where a guide model generates strategies for a core model. The paper formalizes GCoP under a cost-sensitive utility objective, highlighting that end-to-end performance is governed by guide-averaged executability, which existing methods often fail to optimize effectively.