Self-Evolving Multi-Agent Systems via Textual Backpropagation

ACL ARR 2026 January Submission5937 Authors

Keywords: AI agent, Multi-Agents, Generation, Self-Evolving

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 ($\mathcal{ANN}$), 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. Agentic Neural Network 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 $\mathcal{ANN}$ to create new or specialized agent teams post-training, delivering notable gains in accuracy and adaptability. Across seven benchmark datasets, $\mathcal{ANN}$ surpasses leading multi-agent baselines under the same configurations, showing consistent performance improvements.

Paper Type: Long

Research Area: Dialogue and Interactive Systems

Research Area Keywords: Generation, NLP Applications, Multi-Agents, LLM/AI agents

Contribution Types: NLP engineering experiment

Languages Studied: English

Submission Number: 5937