InfiAgent: Self-Evolving Pyramid Agent Framework for Infinite Scenarios

2025-09-26

AI Summary

Current LLM-based agents rely heavily on manually designed workflows, prompt engineering, and domain-expert tuning, limiting scalability and cost-effectiveness. To address this, we propose a pyramid-shaped multi-agent framework grounded in the unified abstraction “agent-as-tool.” It integrates dual-audit quality assurance, DAG-driven dynamic task decomposition and routing, feedback-guided self-evolving agent graph topology, and a parallel atomic-task execution engine. This design enables automatic task dissection, real-time reconfiguration of agent topologies, and cross-scenario adaptive collaboration. Empirically, our framework outperforms the ADAS framework by 9.9% across multiple benchmarks. Moreover, research papers automatically generated using our system received positive evaluations from reviewers at an IEEE flagship conference, demonstrating its effectiveness and generalizability in authentic scientific research settings.

Technology Category

• Multiagent Systems: Adversarial Agents
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work

Abstract

Large Language Model (LLM) agents have demonstrated remarkable capabilities in organizing and executing complex tasks, and many such agents are now widely used in various application scenarios. However, developing these agents requires carefully designed workflows, carefully crafted prompts, and iterative tuning, which requires LLM techniques and domain-specific expertise. These hand-crafted limitations hinder the scalability and cost-effectiveness of LLM agents across a wide range of industries. To address these challenges, we propose extbf{InfiAgent}, a Pyramid-like DAG-based Multi-Agent Framework that can be applied to extbf{infi}nite scenarios, which introduces several key innovations: a generalized "agent-as-a-tool" mechanism that automatically decomposes complex agents into hierarchical multi-agent systems; a dual-audit mechanism that ensures the quality and stability of task completion; an agent routing function that enables efficient task-agent matching; and an agent self-evolution mechanism that autonomously restructures the agent DAG based on new tasks, poor performance, or optimization opportunities. Furthermore, InfiAgent's atomic task design supports agent parallelism, significantly improving execution efficiency. This framework evolves into a versatile pyramid-like multi-agent system capable of solving a wide range of problems. Evaluations on multiple benchmarks demonstrate that InfiAgent achieves 9.9% higher performance compared to ADAS (similar auto-generated agent framework), while a case study of the AI research assistant InfiHelper shows that it generates scientific papers that have received recognition from human reviewers at top-tier IEEE conferences.

Problem

• Automates complex agent workflow design to reduce manual effort
• Enables scalable multi-agent systems for diverse application scenarios
• Improves execution efficiency through autonomous agent evolution mechanisms

Innovation

• Hierarchical multi-agent system with automatic decomposition
• Dual-audit mechanism ensuring task quality and stability
• Self-evolving agent structure adapting to new scenarios