SymbolLKG: Towards Verifiable Logical Reasoning via Logical Knowledge Graph and Symbolic Solvers
cs.AI, cs.CL
Submitted: 2026-08-27
Updated: 2026-08-27
Comments: 19 pages, 9 figures
License: http://creativecommons.org/licenses/by/4.0/
The gist: Large Language Models (LLMs) have demonstrated remarkable proficiency in natural language understanding, yet they struggle with strict multi-step reasoning, frequently suffering from hallucinations
Terminology
Abstract
Large Language Models (LLMs) have demonstrated remarkable proficiency in natural language understanding, yet they struggle with strict multi-step reasoning, frequently suffering from hallucinations and inconsistency. Existing solutions like Chain-of-Thought (CoT) lack rigorous verification mechanisms, while standard Retrieval-Augmented Generation (RAG) often misses the complex, structural dependencies inherent in logical tasks. To bridge this gap, we propose a Neuro-Symbolic architecture that integrates a Logical Knowledge Graph (LKG) with dynamic solver routing. Specifically, we introduce an ontology-based LKG that treats logical rules and constraints as first-class topological nodes, enabling explicit modeling of dependencies extracted from text. We further design a Logic Router to dynamically dispatch tasks to the optimal symbolic engine, which is supported by a topology-aware hybrid retrieval mechanism. Experimental results on logical reasoning benchmarks demonstrate that our framework significantly outperforms state-of-the-art prompting and RAG baselines, delivering higher accuracy and verifiable reasoning paths.
Sources
- BloombergGPT: A Large Language Model for Finance
- Advancing Reasoning in Large Language Models: Promising Methods and Approaches
- Think-on-Graph: Deep and Responsible Reasoning of Large Language Model on Knowledge Graph
- Reasoning on Graphs: Faithful and Interpretable Large Language Model Reasoning
- From Local to Global: A Graph RAG Approach to Query-Focused Summarization
- LightRAG: Simple and Fast Retrieval-Augmented Generation
- Self-Consistency Improves Chain of Thought Reasoning in Language Models
- Spectral Normalisation for Deep Reinforcement Learning: an Optimisation Perspective
- Large Language Models Cannot Self-Correct Reasoning Yet
- Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Tasks
- A Closer Look at Logical Reasoning with LLMs: The Choice of Tool Matters
- StructGPT: A General Framework for Large Language Model to Reason over Structured Data
- Selection-Inference: Exploiting Large Language Models for Interpretable Logical Reasoning
- M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillation
- Language Models Are Greedy Reasoners: A Systematic Formal Analysis of Chain-of-Thought
- Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Steps
- Unsupervised Dense Information Retrieval with Contrastive Learning
- From RAG to Memory: Non-Parametric Continual Learning for Large Language Models
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems
- VSAL: A Vision Solver with Adaptive Layouts for Graph Property Detection