Search papers, labs, and topics across Lattice.
NeSy-RAG is a novel neuro-symbolic retrieval-augmented generation framework that enhances explainability in question answering by integrating Prolog modules with retrieved text chunks. This approach not only generates semantically meaningful predicates but also incorporates a symbolic knowledge-gap detection mechanism to identify and address missing user-specific context. On the ShARC benchmark, NeSy-RAG significantly outperforms a baseline RAG model, achieving 61.1% accuracy compared to 42.8%, demonstrating its effectiveness in providing transparent and attributable reasoning processes.
Achieving over 60% accuracy in explainable question answering, NeSy-RAG links reasoning steps directly to their evidence sources, revolutionizing transparency in LLM outputs.
Retrieval-augmented generation (RAG) improves question answering by grounding large language models (LLMs) in external knowledge such as text corpora. However, its reasoning process remains largely opaque: intermediate reasoning steps are difficult to verify and cannot be reliably attributed to specific evidence. Moreover, missing user-specific context is rarely detected systematically, often leading to incomplete or incorrect output. We propose NeSy-RAG, a modular neuro-symbolic RAG framework that synthesizes attributable Prolog modules from retrieved text chunks. For each chunk, the system generates semantically meaningful predicates that encode Boolean claims, which may depend on user facts. Using joint natural language-code embeddings, predicates are retrieved and composed into Prolog queries. To address incomplete user context, we introduce a symbolic knowledge-gap detection mechanism that identifies missing user facts whose truth values affect the query outcome and automatically triggers follow-up interactions. Executing the resulting Prolog queries yields deterministic answers together with transparent execution traces that link each reasoning step to its originating source. On the ShARC benchmark, without domain-specific training, NeSy-RAG achieves 61.1% accuracy, outperforming a same-model RAG baseline that achieves 42.8% accuracy.
