inference-time-control

Introduces Stochastic Token Steering (STS) and Stochastic Block Steering (SBS) for LLM activation steering, which probabilistically gate steering signals per token or per sequence. Shows that steering only 50% of tokens recovers most of the dense-steering effect while preserving fluency, and that the behavioral outcome is rate-limited by cumulative signal dosage.

This paper introduces a two-stage inference-time budget control method for LLM search agents, using Value-of-Information scores to optimize tool-call and token allocation during multi-hop question answering.