Test-time Prompt Intervention

TL;DR

This paper introduces Test-time Prompt Intervention (PI), a framework that dynamically guides large language models' reasoning during inference, reducing redundancy and hallucinations for more reliable outputs.

Contribution

PI enables real-time, flexible intervention in LLM reasoning processes, integrating human expertise and cognitive principles to improve reasoning quality.

Findings

PI shortens reasoning chains significantly.

PI reduces hallucination in model outputs.

PI improves reasoning reliability and interpretability.

Abstract

Test-time compute has led to remarkable success in the large language model (LLM) community, particularly for complex tasks, where longer chains of thought (CoTs) are generated to enhance reasoning capabilities. However, growing evidence reveals that such reasoning models often produce CoTs plagued by excessive redundancy, including unnecessary verification steps and repetitive reasoning shifts. The root cause lies in post-training of them that overly rely on outcome reward paradigms, as the data of process reward paradigms, which regulate intermediate reasoning steps, is difficult to construct at scale. To address this, we propose PI, a novel framework for Test-time Prompt Intervention. PI provides an interface to dynamically guide and regulate reasoning paths during inference through timely (When module) and proper (How module) interventions and post-intervention sampling (Which module). This allows human problem-solving expertise and cognitive science principles to be seamlessly integrated into LLMs' reasoning processes, enhancing controllability and interpretability. Extensive experiments across multiple models and datasets demonstrate that PI significantly shortens CoTs while reducing hallucination, yielding more concise and reliable reasoning.