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

ICLROral2026

Authors: 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 F. Xu, Shuyan Zhou, Xingyao Wang, Xiang Yue, Tao Yu, Huan Sun, Yu Su, Graham Neubig
Affiliation: Carnegie Mellon University
Venue: ICLR 2026
Track: Oral

TL;DR

We propose Agent Data Protocol (ADP), a lightweight "interlingua" schema that standardizes heterogeneous agent trajectories so datasets can plug into multiple agent SFT pipelines without per-dataset engineering.

Opening excerpt from the authors’ abstract. source

Read the paper

OpenReview

Topics

fine-tuning dataset agent llm

← All ICLR 2026 Oral papers · Browse the whole archive