LogiCity: Advancing Neuro-Symbolic AI with Abstract Urban Simulation

TL;DR

LogiCity presents the first customizable, FOL-grounded urban simulator tailored for Neuro-Symbolic AI, enabling long-horizon multi-agent reasoning with configurable concepts, rules, and agent compositions. It introduces two tasks—Safe Path Following and Visual Action Prediction—to probe abstract reasoning under both sequential and perceptual-noise conditions, using SMT-based grounding and RGB rendering via GPT-4 and diffusion models. The framework demonstrates that NeSy methods can improve compositional generalization for abstract reasoning, while also highlighting challenges in learning complex abstractions, especially in long horizons and high-dimensional data. By providing open-source code and datasets, LogiCity offers a practical platform to design, validate, and compare next-generation NeSy approaches that integrate symbolic reasoning with deep learning in realistic, multi-agent urban environments.

Abstract

Recent years have witnessed the rapid development of Neuro-Symbolic (NeSy) AI systems, which integrate symbolic reasoning into deep neural networks. However, most of the existing benchmarks for NeSy AI fail to provide long-horizon reasoning tasks with complex multi-agent interactions. Furthermore, they are usually constrained by fixed and simplistic logical rules over limited entities, making them far from real-world complexities. To address these crucial gaps, we introduce LogiCity, the first simulator based on customizable first-order logic (FOL) for an urban-like environment with multiple dynamic agents. LogiCity models diverse urban elements using semantic and spatial concepts, such as IsAmbulance(X) and IsClose(X, Y). These concepts are used to define FOL rules that govern the behavior of various agents. Since the concepts and rules are abstractions, they can be universally applied to cities with any agent compositions, facilitating the instantiation of diverse scenarios. Besides, a key feature of LogiCity is its support for user-configurable abstractions, enabling customizable simulation complexities for logical reasoning. To explore various aspects of NeSy AI, LogiCity introduces two tasks, one features long-horizon sequential decision-making, and the other focuses on one-step visual reasoning, varying in difficulty and agent behaviors. Our extensive evaluation reveals the advantage of NeSy frameworks in abstract reasoning. Moreover, we highlight the significant challenges of handling more complex abstractions in long-horizon multi-agent scenarios or under high-dimensional, imbalanced data. With its flexible design, various features, and newly raised challenges, we believe LogiCity represents a pivotal step forward in advancing the next generation of NeSy AI. All the code and data are open-sourced at our website: https://jaraxxus-me.github.io/LogiCity/

Figures (10)

• Figure 1: LogiCity employs abstract concepts and rules, allowing different agent sets to address compositional generalization. Its modular structure enables users to modify abstractions flexibly.
• Figure 2: Demonstration of the Safe Path Following (SPF) and Visual Action Prediction (VAP) tasks in LogiCity. SPF emphasizes sequential decision-making while VAP focuses on one-step sophisticated reasoning on RGB inputs. In the VAP task, we also display the baseline model structure.
• Figure 3: Results for different agent configurations in medium and hard modes of SPF task. We report the average of DSR and TSR here.
• Figure 4: Continual learning results for MLP and NLM dong2019nlm. The results achieved by training from scratch are reported in dashed lines.
• Figure 5: Diverse renderings from LogiCity. Note that every city has distinct agent compositions.