Hengqi Guo
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Hengqi Guo has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Jirui Yang
Fudan University
1 shared publication - Qiang Duan
Pennsylvania State University
1 shared publication - Shijing Hu
Fudan University
1 shared publication - Tao Wei
Ant Group
1 shared publication - Yi Zhao
Fudan University
1 shared publication - Yinggui Wang
Ant Group
1 shared publication
Latest Publications
- Open Access
Prefix Probing: Lightweight Harmful Content Detection for Large Language Models
Large language models often face a three-way trade-off among detection accuracy, inference latency, and deployment cost when used in real-world safety-sensitive applications. This paper introduces Prefix Probing, a black-box harmful content detection method that compares the conditional log-probabilities of "agreement/execution" versus "refusal/safety" opening prefixes and leverages prefix caching to reduce detection overhead to near first-token latency. During inference, the method requires only a single log-probability computation over the probe prefixes to produce a harmfulness score and apply a threshold, without invoking any additional models or multi-stage inference. To further enhance the discriminative power of the prefixes, we design an efficient prefix construction algorithm that automatically discovers highly informative prefixes, substantially improving detection performance. Extensive experiments demonstrate that Prefix Probing achieves detection effectiveness comparable to mainstream external safety models while incurring only minimal computational cost and requiring no extra model deployment, highlighting its strong practicality and efficiency.
Open AccessPrefix Probing: Lightweight Harmful Content Detection for Large Language Models
Large language models often face a three-way trade-off among detection accuracy, inference latency, and deployment cost when used in real-world safety-sensitive applications. This paper introduces Prefix Probing, a black-box harmful content detection method that compares the conditional log-probabilities of "agreement/execution" versus "refusal/safety" opening prefixes and leverages prefix caching to reduce detection overhead to near first-token latency. During inference, the method requires only a single log-probability computation over the probe prefixes to produce a harmfulness score and apply a threshold, without invoking any additional models or multi-stage inference. To further enhance the discriminative power of the prefixes, we design an efficient prefix construction algorithm that automatically discovers highly informative prefixes, substantially improving detection performance. Extensive experiments demonstrate that Prefix Probing achieves detection effectiveness comparable to mainstream external safety models while incurring only minimal computational cost and requiring no extra model deployment, highlighting its strong practicality and efficiency.