Papperoni

2026

Agent Data Protocol: Unifying Datasets for Diverse, Effective Fine-Tuning of LLM Agents

Yueqi Song, Ketan Ramaneti, Zaid Sheikh, Ziru Chen, Boyu Gou, Tianbao Xie, Yiheng Xu, Danyang Zhang, Apurva Gandhi, Fan Yang, Joseph Liu, Tianyue Ou, Zhihao Yuan, Frank Xu, Shuyan Zhou, Xiang Yue, Tao Yu, Huan Sun, Yu Su, Graham Neubig

Open PDF Google Scholar

citations

Cite Score

AI summary

This paper introduces the Agent Data Protocol (ADP), a light-weight representation language, to unify 13 existing agent training datasets, achieving an average performance gain of ~20% over base models and state-of-the-art or near-SOTA results on coding, browsing, tool use, and research benchmarks without domain-specific tuning.

Main Contributions

Introduction of the Agent Data Protocol (ADP), a standardized, expressive representation language for agent data.
Unification of 13 diverse existing agent training datasets into the ADP format, creating the largest publicly available dataset for agent training (ADP Dataset V1) with 1.3M trajectories.
Demonstration of significant performance improvements (average ~20% gain over base models) through supervised fine-tuning on ADP-unified data across various domains.
Achievement of state-of-the-art or near-SOTA performance on coding, browsing, tool use, and research benchmarks using ADP, without domain-specific tuning.
Release of all code and datasets in open source to promote standardized, scalable, and reproducible agent training.

Abstract

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 supervised finetuning on the unified 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.

Citation Graph

Loading graph...

References [58]

Sort:

Filter:

[1]SQuAD: 100,000+ Questions for Machine Comprehension of Text

P. Rajpurkar, J. Zhang, K. Lopyrev, Percy Liang - 2016

37 papers in library cite