Hephaestus: Improving Fundamental Agent Capabilities of Large Language Models through Continual Pre-Training
Yuchen Zhuang, Jingfeng Yang, Haoming Jiang, Xin Liu, Kewei Cheng, Sanket Lokegaonkar, Yifan Gao, Qing Ping
Abstract
Due to the scarcity of agent-oriented pre-training data, LLM-based autonomous agents typically rely on complex prompting or extensive fine-tuning, which often fails to introduce new capabilities while preserving strong generalizability. We introduce Hephaestus-Forge, the first large-scale pre-training corpus designed to enhance the fundamental capabilities of LLM agents in API function calling, intrinsic reasoning and planning, and adapting to environmental feedback. Hephaestus-Forge comprises 103B agent-specific data encompassing 76,537 APIs, including both tool documentation to introduce knowledge of API functions and function calling trajectories to strengthen intrinsic reasoning. To explore effective training protocols, we investigate scaling laws to identify the optimal recipe in data mixing ratios. By continual pre-training on Hephaestus-Forge, Hephaestus outperforms small- to medium-scale open-source LLMs and rivals commercial LLMs on three agent benchmarks, demonstrating the effectiveness of our pre-training corpus in enhancing fundamental agentic capabilities and generalization of LLMs to new tasks or environments.
BibTeX
@inproceedings{zhuang-etal-2025-hephaestus,
title = "Hephaestus: Improving Fundamental Agent Capabilities of Large Language Models through Continual Pre-Training",
author = "Zhuang, Yuchen and
Yang, Jingfeng and
Jiang, Haoming and
Liu, Xin and
Cheng, Kewei and
Lokegaonkar, Sanket and
Gao, Yifan and
Ping, Qing and
Liu, Tianyi and
Huang, Binxuan and
Li, Zheng and
Wang, Zhengyang and
Chen, Pei and
Wang, Ruijie and
Zhang, Rongzhi and
Zalmout, Nasser and
Nigam, Priyanka and
Yin, Bing and
Zhang, Chao",
editor = "Chiruzzo, Luis and
Ritter, Alan and
Wang, Lu",
booktitle = "Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)",
month = apr,
year = "2025",
address = "Albuquerque, New Mexico",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.naacl-long.308/",
pages = "6041--6068",
ISBN = "979-8-89176-189-6"
}