LogicGuard: Improving embodied LLM agents through temporal logic based critics

Anand Gokhale, Vaibhav Srivastava, Francesco Bullo

Keywords: LLM, Embodied Agents, Planning, Safety, Formal Logic

TL;DR: We propose an actor–critic framework where LLMs supervise each other, communicating via temporal logic, improving task completion, efficiency, and safety across general short horizon and specialized long-horizon planning tasks.

Abstract: Large language models (LLMs) have shown promise in zero-shot and single step reasoning and decision-making problems, but in long-horizon sequential planning tasks, their errors compound, often leading to unreliable or inefficient behavior. We introduce LogicGuard, a modular actor–critic architecture in which an LLM actor is guided by a trajectory-level LLM critic that communicates through Linear Temporal Logic (LTL). Our setup combines the reasoning strengths of language models with the guarantees of formal logic. The actor selects high-level actions from natural language observations, while the critic analyzes full trajectories and proposes new LTL constraints that shield the actor from future unsafe or inefficient behavior. LogicGuard supports both fixed safety rules and adaptive, learned constraints, and is model-agnostic: any LLM-based planner can serve as the actor, with LogicGuard acting as a logic-generating wrapper. We formalize planning as graph traversal under symbolic constraints, allowing LogicGuard to analyze failed or suboptimal trajectories and generate new temporal logic rules that improve future behavior. To demonstrate generality, we evaluate LogicGuard across two distinct settings: short-horizon general tasks and long-horizon specialist tasks. On the Behavior benchmark of 100 household tasks, LogicGuard increases task completion rates by 25\% over a baseline InnerMonologue planner. On the Minecraft diamond-mining task, which is long-horizon and requires multiple interdependent subgoals, LogicGuard improves both efficiency and safety compared to SayCan and InnerMonologue. These results show that enabling LLMs to supervise each other through temporal logic yields more reliable, efficient and safe decision-making for both embodied agents.

Primary Area: applications to robotics, autonomy, planning

Submission Number: 9449