Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Agentic Tool Use in Large Language Models

Jinchao Hu, Meizhi Zhong, Kehai Chen, Xuefeng Bai, Min Zhang

Abstract

Large language models are increasingly being deployed as autonomous agents yet their real world effectiveness depends on reliable tools for information retrieval, computation and external action. Existing studies remain fragmented across tasks, tool types, and training settings, lacking a unified view of how tool-use methods differ and evolve. This paper organizes the literature into three paradigms: prompting as plug-and-play, supervised tool learning and reward-driven tool policy learning, analyzes their methods, strengths and failure modes, reviews the evaluation landscape and highlights key challenges, aiming to address this fragmentation and provide a more structured evolutionary view of agentic tool use.

Agentic Tool Use in Large Language Models

Abstract

Large language models are increasingly being deployed as autonomous agents yet their real world effectiveness depends on reliable tools for information retrieval, computation and external action. Existing studies remain fragmented across tasks, tool types, and training settings, lacking a unified view of how tool-use methods differ and evolve. This paper organizes the literature into three paradigms: prompting as plug-and-play, supervised tool learning and reward-driven tool policy learning, analyzes their methods, strengths and failure modes, reviews the evaluation landscape and highlights key challenges, aiming to address this fragmentation and provide a more structured evolutionary view of agentic tool use.

Paper Structure

This paper contains 25 sections, 3 equations, 5 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Background of Agent Tool Use
  3. Early Exploration and Concept Formation
  4. The Emergence of Prompt-Driven Agency (2022-2023)
  5. The Shift to Supervised Tool Learning (2023-2024)
  6. The Rise of Policy for Autonomous Optimization (2024-Present)
  7. Prompting as plug-and-play
  8. Interleaved Reasoning and Action
  9. Decoupled Planning and Execution
  10. Program-Aided Reasoning
  11. Supervised Tool Learning
  12. Self-Supervised Data Generation
  13. Large-Scale Instruction Tuning
  14. Process-Oriented and Alignment-Focused Tuning
  15. Reward-Driven Tool Policy Learning
  16. ...and 10 more sections

Figures (5)

  • Figure 1: Overview of the Agentic Tool Use framework.
  • Figure 2: Timeline of Agentic Tool Use.
  • Figure 3: Illustration of the Prompting as plug-and-play paradigm.
  • Figure 4: Training Pipeline of Paradigm II: From Supervision to Internalized Tool-Use Capability.
  • Figure 5: An overview of Paradigm III: Reward-Driven Tool Policy Learning.