Training-Free Agentic AI: Probabilistic Control and Coordination in Multi-Agent LLM Systems

TL;DR

This paper presents REDEREF, a training-free probabilistic controller for multi-agent LLM systems that enhances routing efficiency, robustness, and reduces resource usage through belief-guided delegation and reflection-driven re-routing.

Contribution

The paper introduces REDEREF, a novel, training-free control method that improves multi-agent LLM collaboration by integrating probabilistic routing, reflection, and memory-aware priors.

Findings

Reduces token usage by 28%

Decreases agent calls by 17%

Cuts time-to-success by 19%

Abstract

Multi-agent large language model (LLM) systems enable complex, long-horizon reasoning by composing specialized agents, but practical deployment remains hindered by inefficient routing, noisy feedback, and high interaction cost. We introduce REDEREF, a lightweight and training-free controller for multi-agent LLM collaboration that improves routing efficiency during recursive delegation. REDEREF integrates (i) belief-guided delegation via Thompson sampling to prioritize agents with historically positive marginal contributions, (ii) reflection-driven re-routing using a calibrated LLM or programmatic judge, (iii) evidence-based selection rather than output averaging, and (iv) memory-aware priors to reduce cold-start inefficiency. Across multi-agent split-knowledge tasks, we show that while recursive retry alone saturates task success, belief-guided routing reduces token usage by 28%, agent calls by 17%, and time-to-success by 19% compared to random recursive delegation, and adapts gracefully under agent or judge degradation. These results demonstrate that simple, interpretable probabilistic control can meaningfully improve the efficiency and robustness of multi-agent LLM systems without training or fine-tuning.