OmniReflect: Discovering Transferable Constitutions for LLM agents via Neuro-Symbolic Reflections

TL;DR

OmniReflect introduces a neuro-symbolic, reflection-based framework that creates a set of guiding principles to significantly improve the performance and efficiency of LLM agents across various complex tasks and environments.

Contribution

This work presents OmniReflect, a novel hierarchical framework that constructs transferable constitutions for LLM agents using neuro-symbolic techniques, enabling better long-term learning and adaptability.

Findings

Major performance improvements across multiple benchmarks

Effective in both self-sustaining and co-operative modes

Outperforms existing baselines with significant success rate gains

Abstract

Efforts to improve Large Language Model (LLM) agent performance on complex tasks have largely focused on fine-tuning and iterative self-correction. However, these approaches often lack generalizable mechanisms for longterm learning and remain inefficient in dynamic environments. We introduce OmniReflect, a hierarchical, reflection-driven framework that constructs a constitution, a compact set of guiding principles distilled from task experiences, to enhance the effectiveness and efficiency of an LLM agent. OmniReflect operates in two modes: Self-sustaining, where a single agent periodically curates its own reflections during task execution, and Co-operative, where a Meta-advisor derives a constitution from a small calibration set to guide another agent. To construct these constitutional principles, we employ Neural, Symbolic, and NeuroSymbolic techniques, offering a balance between contextual adaptability and computational efficiency. Empirical results averaged across models show major improvements in task success, with absolute gains of +10.3% on ALFWorld, +23.8% on BabyAI, and +8.3% on PDDL in the Self-sustaining mode. Similar gains are seen in the Co-operative mode, where a lightweight Qwen3-4B ReAct agent outperforms all Reflexion baselines on BabyAI. These findings highlight the robustness and effectiveness of OmniReflect across environments and backbones.