REASON: Accelerating Probabilistic Logical Reasoning for Scalable Neuro-Symbolic Intelligence

2026-01-28

AI Summary

This work addresses the inefficiency of probabilistic logical reasoning in neuro-symbolic AI on conventional CPU/GPU hardware, which stems from irregular control flow, low arithmetic intensity, and sparse memory access. To overcome these challenges, the authors propose REASON, a novel framework that unifies probabilistic logical reasoning as a directed acyclic graph and introduces a reconfigurable tree-based processing architecture tailored for irregular traversals and fused symbolic-probabilistic computation. By integrating adaptive pruning, multi-level pipelining, and deep co-design with GPU streaming multiprocessors, REASON achieves substantial performance and energy efficiency gains. Evaluated across six neuro-symbolic workloads, it delivers 12–50× speedup and 310–681× improvement in energy efficiency, completing end-to-end inference in just 0.8 seconds within a 6 mm² chip area and 2.12 W power consumption.

Technology Category

• Knowledge Representation and Reasoning: Computational Complexity of Reasoning
• Machine Learning: Probabilistic Circuits and Graphical Models
• Reasoning under Uncertainty: Probabilistic Programming

Application Category

• Search and Retrieval-Augmented AI: Large language models for search
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Neuro-symbolic AI systems integrate neural perception with symbolic reasoning to enable data-efficient, interpretable, and robust intelligence beyond purely neural models. Although this compositional paradigm has shown superior performance in domains such as reasoning, planning, and verification, its deployment remains challenging due to severe inefficiencies in symbolic and probabilistic inference. Through systematic analysis of representative neuro-symbolic workloads, we identify probabilistic logical reasoning as the inefficiency bottleneck, characterized by irregular control flow, low arithmetic intensity, uncoalesced memory accesses, and poor hardware utilization on CPUs and GPUs. This paper presents REASON, an integrated acceleration framework for probabilistic logical reasoning in neuro-symbolic AI. REASON introduces a unified directed acyclic graph representation that captures common structure across symbolic and probabilistic models, coupled with adaptive pruning and regularization. At the architecture level, REASON features a reconfigurable, tree-based processing fabric optimized for irregular traversal, symbolic deduction, and probabilistic aggregation. At the system level, REASON is tightly integrated with GPU streaming multiprocessors through a programmable interface and multi-level pipeline that efficiently orchestrates compositional execution. Evaluated across six neuro-symbolic workloads, REASON achieves 12-50x speedup and 310-681x energy efficiency over desktop and edge GPUs under TSMC 28 nm node. REASON enables real-time probabilistic logical reasoning, completing end-to-end tasks in 0.8 s with 6 mm2 area and 2.12 W power, demonstrating that targeted acceleration of probabilistic logical reasoning is critical for practical and scalable neuro-symbolic AI and positioning REASON as a foundational system architecture for next-generation cognitive intelligence.

Problem

• probabilistic logical reasoning
• neuro-symbolic AI
• inference inefficiency
• hardware utilization
• scalability

Innovation

• probabilistic logical reasoning
• neuro-symbolic AI
• hardware acceleration
• reconfigurable architecture
• directed acyclic graph