RELIC: Revealed Principles for Learning Interpretable Composable Skills in Multi-Agent Planning
Artificial IntelligenceMultiagent Systems
Summary
The authors address the challenge of multiple agents working together while keeping their internal decision-making private. They developed RELIC, a system where each agent improves its skills independently using large language models, and a trusted coordinator checks only the overall team's success. Instead of sharing raw code, agents communicate successful strategies as simple principles that others can adapt locally. This approach helps different agents cooperate without exposing how they work internally.
Authors
Nguyen Viet Tuan Kiet, Bui Dinh Pham, Duong Quoc Chinh, Dao Van Tung, Tran Cong Dao, Huynh Thi Thanh Binh
Abstract
Multi-agent planning becomes substantially harder when agents must improve specialized decision-making skills while keeping their internal implementations private. This regime arises when agents are developed independently, expose different interfaces and capabilities, and must nevertheless coordinate without sharing executable policies. Prior research has largely assumed centralized optimization, shared policy access, or common skill representations, making it poorly suited to privacy-constrained cooperation. We introduce RELIC, a framework for learning interpretable and composable skills via revealed principles. Each agent refines its own programmatic skill through private LLM-guided search, while a trusted orchestrator evaluates proposed updates solely through team-level performance. Successful behaviors are not broadcast as code; instead, they are abstracted into portable principles that other agents can instantiate within their own interfaces and recombine with local strategies. This separates coordination from implementation sharing, enabling cross-agent transfer under heterogeneous skill signatures. RELIC thus introduces a new paradigm for privacy-preserving skill learning and coordination in multi-agent planning.