AgentCPM-Explore: Realizing Long-Horizon Deep Exploration for Edge-Scale Agents
Abstract
While Large Language Model (LLM)-based agents have shown remarkable potential for solving complex tasks, existing systems remain heavily reliant on large-scale models, leaving the capabilities of edge-scale models largely underexplored. In this paper, we present the first systematic study on training agentic models at the 4B-parameter scale. We identify three primary bottlenecks hindering the performance of edge-scale models: catastrophic forgetting during Supervised Fine-Tuning (SFT), sensitivity to reward signal noise during Reinforcement Learning (RL), and reasoning degradation caused by redundant information in long-context scenarios. To address the issues, we propose AgentCPM-Explore, a compact 4B agent model with high knowledge density and strong exploration capability. We introduce a holistic training framework featuring parameter-space model fusion, reward signal denoising, and contextual information refinement. Through deep exploration, AgentCPM-Explore achieves state-of-the-art (SOTA) performance among 4B-class models, matches or surpasses 8B-class SOTA models on four benchmarks, and even outperforms larger-scale models such as Claude-4.5-Sonnet or DeepSeek-v3.2 in five benchmarks. Notably, AgentCPM-Explore achieves 97.09% accuracy on GAIA text-based tasks under pass@64. These results provide compelling evidence that the bottleneck for edge-scale models is not their inherent capability ceiling, but rather their inference stability. Based on our well-established training framework, AgentCPM-Explore effectively unlocks the significant, yet previously underestimated, potential of edge-scale models.
Keywords
Cite
@article{arxiv.2602.06485,
title = {AgentCPM-Explore: Realizing Long-Horizon Deep Exploration for Edge-Scale Agents},
author = {Haotian Chen and Xin Cong and Shengda Fan and Yuyang Fu and Ziqin Gong and Yaxi Lu and Yishan Li and Boye Niu and Chengjun Pan and Zijun Song and Huadong Wang and Yesai Wu and Yueying Wu and Zihao Xie and Yukun Yan and Zhong Zhang and Yankai Lin and Zhiyuan Liu and Maosong Sun},
journal= {arXiv preprint arXiv:2602.06485},
year = {2026}
}