AgentTuning: Enabling Generalized Agent Abilities for LLMs

By Aohan Zeng, Mingdao Liu, Rui Lu, Bowen Wang, Xiao Liu, Yuxiao Dong, Jie Tang. Annual Meeting of the Association for Computational Linguistics, 2023

347 citations recorded by the source corpus.

Open large language models (LLMs) with great performance in various tasks have significantly advanced the development of LLMs. However, they are far inferior to commercial models such as ChatGPT and GPT-4 when acting as agents to tackle complex tasks in the real world. These agent tasks employ LLMs as the central controller responsible for planning, memorization, and tool utilization, necessitating both fine-grained prompting methods and robust LLMs to achieve satisfactory performance. Though many prompting methods have been proposed to complete particular agent tasks, there is lack of research focusing on improving the agent capabilities of LLMs themselves without compromising their general abilities. In this work, we present AgentTuning, a simple and general method to enhance the agent abilities of LLMs while maintaining their general LLM capabilities. We construct AgentInstruct, a lightweight instruction-tuning dataset containing high-quality interaction trajectories. We employ a hybrid instruction-tuning strategy by combining AgentInstruct with open-source instructions from general domains. AgentTuning is used to instruction-tune the Llama 2 series, resulting in AgentLM. Our evaluations show that AgentTuning enables LLMs' agent capabilities without compromising general abilities. The AgentLM-70B is comparable to GPT-3.5-turbo on unseen agent tasks, demonstrating generalized agent capabilities. We open source the AgentInstruct and AgentLM-7B, 13B, and 70B models at https://github.com/THUDM/AgentTuning, serving open and powerful alternatives to commercial LLMs for agent tasks.

Subfields: Artificial Intelligence, Health Informatics

View Paper: https://www.semanticscholar.org/paper/46fe9ce789408b8a50fb4259e6bf0cc5855f4ed5

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Graham Neubig
2. Shuyan Zhou
3. Shirui Pan
4. Yufei Wang
5. Xindong Wu
6. Linhao Luo
7. Gao Huang
8. Robert Lo

Citations in the corpus

• ExpeL: LLM Agents Are Experiential Learners (2023, background): “Exploring LLM agents us-ing open-source LLMs should be another promising future work (Zeng et al. 2023).”
• CogAgent: A Visual Language Model for GUI Agents (2023, neutral): “For instance, AgentTuning [42] and MindAct [10] evaluated Llama2-70B and Flan-T5-XL in a fine-tuned setting, and GPT-3.”
• Agentic AI: Autonomous Intelligence for Complex Goals—A Comprehensive Survey (2025, neutral): “These include reasoning and planning, tool use, memory mechanisms, Retrieval-Augmented Generation (RAG), and instruction fine-tuning [23].”
• Unifying Large Language Models and Knowledge Graphs: A Roadmap (2023, methodology): “To enhance the agent abilities, AgentTuning [248] presents several instruction-tuning datasets to guide LLM agents to perform reasoning on KGs. Comparison and Discussion.”
• VisualWebArena: Evaluating Multimodal Agents on Realistic Visual Web Tasks (2024, background): “Zeng et al. (2023) fine-tunes the LLaMA-2 (Touvron et al., 2023b) models on a set of interaction trajectories with instructions, improving over baseline agent models.”