Enhancing Large Language Models with Neurosymbolic Reasoning for Multilingual Tasks

TL;DR

This paper introduces NeuroSymbolic Augmented Reasoning (NSAR), a novel approach that combines neural and symbolic reasoning to improve multilingual multi-target reasoning in large language models, especially in long-context scenarios.

Contribution

The paper presents NSAR, a new method that explicitly extracts symbolic facts and generates executable code, enhancing reasoning accuracy and interpretability in multilingual tasks.

Findings

NSAR outperforms vanilla RAG and prompting strategies across seven languages.

NSAR improves reasoning accuracy in long-context scenarios.

The approach offers scalable and interpretable reasoning in multilingual settings.

Abstract

Large language models (LLMs) often struggle to perform multi-target reasoning in long-context scenarios where relevant information is scattered across extensive documents. To address this challenge, we introduce NeuroSymbolic Augmented Reasoning (NSAR), which combines the benefits of neural and symbolic reasoning during inference. NSAR explicitly extracts symbolic facts from text and generates executable Python code to handle complex reasoning steps. Through extensive experiments across seven languages and diverse context lengths, we demonstrate that NSAR significantly outperforms both a vanilla RAG baseline and advanced prompting strategies in accurately identifying and synthesizing multiple pieces of information. Our results highlight the effectiveness of combining explicit symbolic operations with neural inference for robust, interpretable, and scalable reasoning in multilingual settings.