Scientific Knowledge-driven Decoding Constraints Improving the Reliability of LLMs

TL;DR

SciDC enhances LLM reliability by integrating scientific knowledge and constraints, significantly reducing hallucinations and improving accuracy across scientific tasks.

Contribution

This paper introduces SciDC, a novel framework that converts scientific knowledge into constraints to improve LLM generation on domain-specific tasks.

Findings

Achieved 12% average accuracy improvement over vanilla models.

Effectively constrains model generation on scientific tasks.

Demonstrated potential for automatic knowledge summarization.

Abstract

Large language models (LLMs) have shown strong knowledge reserves and task-solving capabilities, but still face the challenge of severe hallucination, hindering their practical application. Though scientific theories and rules can efficiently direct the behaviors of human manipulators, LLMs still do not utilize these highly-condensed knowledge sufficiently through training or prompting. To address this issue, we propose \textbf{SciDC}, an LLM generation method that integrate subject-specific knowledge with strong constraints. By adopting strong LLMs to automatically convert flexible knowledge into multi-layered, standardized rules, we build an extensible framework to effectively constrain the model generation on domain tasks. Experiments on scientific tasks including industrial formulation design, clinical tumor diagnosis and retrosynthesis planning, consistently demonstrate the effectiveness of our method, achieving a 12\% accuracy improvement on average compared with vanilla generation. We further discuss the potential of LLMs in automatically inductively summarizing highly-condensed knowledge, looking ahead to practical solutions for accelerating the overall scientific research process. All the code of this paper can be obtained (https://github.com/Maotian-Ma/SciDC).