Self-Evolving Multi-Agent Systems via Textual Backpropagation

[Submitted on 10 Jun 2025 (v1), last revised 16 Jun 2026 (this version, v3)]

View PDF HTML (experimental)

Abstract:Leveraging multiple Large Language Models (LLMs) has proven effective for addressing complex, high-dimensional tasks, but current approaches often rely on static, manually engineered multi-agent configurations. To overcome these constraints, we present the Agentic Neural Network (ANN), a framework that conceptualizes multi-agent collaboration as a layered neural network architecture. In this design, each agent operates as a node, and each layer forms a cooperative team focused on a specific subtask. Our framework follows a two-phase optimization strategy: (1) Forward Phase – Drawing inspiration from neural network forward passes, tasks are dynamically decomposed into subtasks, and cooperative agent teams with suitable aggregation methods are constructed layer by layer. (2) Backward Phase – Mirroring backpropagation, we refine both global and local collaboration through iterative feedback, allowing agents to self-evolve their roles, prompts, and coordination. This neuro-symbolic approach enables our framework to create new or specialized agent teams post-training, delivering notable gains in accuracy and adaptability. Across seven benchmark datasets, our work surpasses leading multi-agent baselines under the same configurations, showing consistent performance improvements.

Submission history

From: Xiaowen Ma [view email]
[v1] Tue, 10 Jun 2025 17:59:21 UTC (4,614 KB)
[v2] Fri, 18 Jul 2025 14:52:19 UTC (4,621 KB)
[v3] Tue, 16 Jun 2026 19:18:36 UTC (5,379 KB)