AgentCPM-Explore: Realizing Long-Horizon Deep Exploration for Edge-Scale Agents
About
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.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| General AI Assistant Task Completion | GAIA Text-Only | Accuracy0.639 | 15 | |
| Deep Information Search and Synthesis | xbench DeepSearch | Score70 | 14 | |
| Web Browsing Competition | Browse Comp | Score24.1 | 14 | |
| Expert-Level Question Answering | Humanity's Last Exam | Accuracy19.1 | 14 | |
| Web Navigation Question Answering | WebWalker QA | Accuracy68.1 | 13 | |
| Web Browsing Competition (Chinese) | Browse Comp ZH | Score29.1 | 13 | |
| Fact Retrieval and Analysis | FRAMES | Accuracy82.7 | 9 | |
| Agent Capability Evaluation | SEAL 0 | Score40.5 | 9 |