AgentCPM-Explore: Realizing Long-Horizon Deep Exploration for Edge-Scale Agents

• URL: http://arxiv.org/abs/2602.06485v1
• Date: Fri, 06 Feb 2026 08:24:59 GMT
• Title: AgentCPM-Explore: Realizing Long-Horizon Deep Exploration for Edge-Scale Agents
• Authors: Haotian Chen, Xin Cong, Shengda Fan, Yuyang Fu, Ziqin Gong, Yaxi Lu, Yishan Li, Boye Niu, Chengjun Pan, Zijun Song, Huadong Wang, Yesai Wu, Yueying Wu, Zihao Xie, Yukun Yan, Zhong Zhang, Yankai Lin, Zhiyuan Liu, Maosong Sun,
• Abstract summary: AgentCPM-Explore is a compact 4B agent model with high knowledge density and strong exploration capability.<n>We introduce a holistic training framework featuring parameter-space model fusion, reward signal denoising, and contextual information refinement.<n>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.
• Score: 75.67445299298949
• License: http://creativecommons.org/licenses/by/4.0/
• 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.

Related papers

• Youtu-LLM: Unlocking the Native Agentic Potential for Lightweight Large Language Models [78.73992315826035]
  We introduce Youtu-LLM, a lightweight language model that harmonizes high computational efficiency with native agentic intelligence.<n>Youtu-LLM is pre-trained from scratch to systematically cultivate reasoning and planning capabilities.
  arXiv  Detail & Related papers  (2025-12-31T04:25:11Z)
• How Do LLMs Fail In Agentic Scenarios? A Qualitative Analysis of Success and Failure Scenarios of Various LLMs in Agentic Simulations [0.0]
  We investigate how large language models (LLMs) fail when operating as autonomous agents with tool-use capabilities.<n>Using the Kamiwaza Agentic Merit Index (KAMI) v0.1 benchmark, we analyze 900 execution traces from three representative models.<n>We identify four recurring failure archetypes: premature action without grounding, over-helpfulness that substitutes missing entities, vulnerability to distractor-induced context pollution, and fragile execution.
  arXiv  Detail & Related papers  (2025-12-08T12:27:15Z)
• Klear-AgentForge: Forging Agentic Intelligence through Posttraining Scaling [46.593200463657645]
  We present a comprehensive and fully open-source pipeline for training a high-performance agentic model, named Klear-Qwen3-AgentForge.<n>We design effective supervised fine-tuning (SFT) with synthetic data followed by multi-turn reinforcement learning (RL) to unlock the potential for multiple diverse agentic tasks.
  arXiv  Detail & Related papers  (2025-11-08T09:47:27Z)
• One Model to Critique Them All: Rewarding Agentic Tool-Use via Efficient Reasoning [54.580646706013965]
  Reward models (RMs) play a critical role in aligning large language models with human preferences.<n>We introduce ToolRM, a family of lightweight generative RMs tailored for general tool-use scenarios.<n>To build these models, we propose a novel pipeline that constructs pairwise preference data using rule-based scoring and multidimensional sampling.
  arXiv  Detail & Related papers  (2025-10-30T06:08:27Z)
• Demystifying Reinforcement Learning in Agentic Reasoning [90.3737088727791]
  We conduct a comprehensive and systematic investigation to demystify reinforcement learning in agentic reasoning.<n>We highlight our key insights: (i) replacing stitched synthetic trajectories with real end-to-end tool-use trajectories yields a far stronger SFT.<n> Exploration-friendly techniques are crucial for agentic RL, such as clip higher, overlong reward shaping, and maintaining adequate policy entropy could improve the training efficiency.
  arXiv  Detail & Related papers  (2025-10-13T17:57:15Z)
• Scheming Ability in LLM-to-LLM Strategic Interactions [4.873362301533824]
  Large language model (LLM) agents are deployed autonomously in diverse contexts.<n>We investigate the ability and propensity of frontier LLM agents through two game-theoretic frameworks.<n>Tests four models (GPT-4o, Gemini-2.5-pro, Claude-3.7-Sonnet, and Llama-3.3-70b)
  arXiv  Detail & Related papers  (2025-10-11T04:42:29Z)
• Scaling Agents via Continual Pre-training [80.97989245493326]
  We propose incorporating Agentic Continual Pre-training (Agentic CPT) into the deep research agents training pipeline to build powerful agentic foundational models.<n>We evaluate our AgentFounder-30B on 10 benchmarks and achieve state-of-the-art performance while retains strong tool-use ability.
  arXiv  Detail & Related papers  (2025-09-16T17:57:19Z)
• Phi-4-Mini-Reasoning: Exploring the Limits of Small Reasoning Language Models in Math [135.1260782461186]
  Chain-of-Thought (CoT) significantly enhances formal reasoning capabilities in Large Language Models (LLMs)<n>However, improving reasoning in Small Language Models (SLMs) remains challenging due to their limited model capacity.<n>We present a systematic training recipe for SLMs that consists of four steps: (1) large-scale mid-training on diverse distilled long-CoT data, (2) supervised fine-tuning on high-quality long-CoT data, (3) Rollout DPO leveraging a carefully curated preference dataset, and (4) Reinforcement Learning (RL) with Verifiable Reward.
  arXiv  Detail & Related papers  (2025-04-30T00:04:35Z)
• CausalAgents: A Robustness Benchmark for Motion Forecasting using Causal Relationships [8.679073301435265]
  We construct a new benchmark for evaluating and improving model robustness by applying perturbations to existing data. We use these labels to perturb the data by deleting non-causal agents from the scene. Under non-causal perturbations, we observe a $25$-$38%$ relative change in minADE as compared to the original.
  arXiv  Detail & Related papers  (2022-07-07T21:28:23Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.