Saved in:
Bibliographic Details
Main Authors: Song, Yueqi, Ramaneti, Ketan, Sheikh, Zaid, Chen, Ziru, Gou, Boyu, Xie, Tianbao, Xu, Yiheng, Zhang, Danyang, Gandhi, Apurva, Yang, Fan, Liu, Joseph, Ou, Tianyue, Yuan, Zhihao, Xu, Frank, Zhou, Shuyan, Wang, Xingyao, Yue, Xiang, Yu, Tao, Sun, Huan, Su, Yu, Neubig, Graham
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2510.24702
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912940810043392
author Song, Yueqi
Ramaneti, Ketan
Sheikh, Zaid
Chen, Ziru
Gou, Boyu
Xie, Tianbao
Xu, Yiheng
Zhang, Danyang
Gandhi, Apurva
Yang, Fan
Liu, Joseph
Ou, Tianyue
Yuan, Zhihao
Xu, Frank
Zhou, Shuyan
Wang, Xingyao
Yue, Xiang
Yu, Tao
Sun, Huan
Su, Yu
Neubig, Graham
author_facet Song, Yueqi
Ramaneti, Ketan
Sheikh, Zaid
Chen, Ziru
Gou, Boyu
Xie, Tianbao
Xu, Yiheng
Zhang, Danyang
Gandhi, Apurva
Yang, Fan
Liu, Joseph
Ou, Tianyue
Yuan, Zhihao
Xu, Frank
Zhou, Shuyan
Wang, Xingyao
Yue, Xiang
Yu, Tao
Sun, Huan
Su, Yu
Neubig, Graham
contents Public research results on large-scale supervised finetuning of AI agents remain relatively rare, since the collection of agent training data presents unique challenges. In this work, we argue that the bottleneck is not a lack of underlying data sources, but that a large variety of data is fragmented across heterogeneous formats, tools, and interfaces. To this end, we introduce the agent data protocol (ADP), a light-weight representation language that serves as an "interlingua" between agent datasets in diverse formats and unified agent training pipelines downstream. The design of ADP is expressive enough to capture a large variety of tasks, including API/tool use, browsing, coding, software engineering, and general agentic workflows, while remaining simple to parse and train on without engineering at a per-dataset level. In experiments, we unified a broad collection of 13 existing agent training datasets into ADP format, and converted the standardized ADP data into training-ready formats for multiple agent frameworks. We performed SFT on these data, and demonstrated an average performance gain of ~20% over corresponding base models, and delivers state-of-the-art or near-SOTA performance on standard coding, browsing, tool use, and research benchmarks, without domain-specific tuning. All code and data are released publicly, in the hope that ADP could help lower the barrier to standardized, scalable, and reproducible agent training.
format Preprint
id arxiv_https___arxiv_org_abs_2510_24702
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Agent Data Protocol: Unifying Datasets for Diverse, Effective Fine-tuning of LLM Agents
Song, Yueqi
Ramaneti, Ketan
Sheikh, Zaid
Chen, Ziru
Gou, Boyu
Xie, Tianbao
Xu, Yiheng
Zhang, Danyang
Gandhi, Apurva
Yang, Fan
Liu, Joseph
Ou, Tianyue
Yuan, Zhihao
Xu, Frank
Zhou, Shuyan
Wang, Xingyao
Yue, Xiang
Yu, Tao
Sun, Huan
Su, Yu
Neubig, Graham
Computation and Language
Artificial Intelligence
Public research results on large-scale supervised finetuning of AI agents remain relatively rare, since the collection of agent training data presents unique challenges. In this work, we argue that the bottleneck is not a lack of underlying data sources, but that a large variety of data is fragmented across heterogeneous formats, tools, and interfaces. To this end, we introduce the agent data protocol (ADP), a light-weight representation language that serves as an "interlingua" between agent datasets in diverse formats and unified agent training pipelines downstream. The design of ADP is expressive enough to capture a large variety of tasks, including API/tool use, browsing, coding, software engineering, and general agentic workflows, while remaining simple to parse and train on without engineering at a per-dataset level. In experiments, we unified a broad collection of 13 existing agent training datasets into ADP format, and converted the standardized ADP data into training-ready formats for multiple agent frameworks. We performed SFT on these data, and demonstrated an average performance gain of ~20% over corresponding base models, and delivers state-of-the-art or near-SOTA performance on standard coding, browsing, tool use, and research benchmarks, without domain-specific tuning. All code and data are released publicly, in the hope that ADP could help lower the barrier to standardized, scalable, and reproducible agent training.
title Agent Data Protocol: Unifying Datasets for Diverse, Effective Fine-tuning of LLM Agents
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2510.24702