Salvato in:
Dettagli Bibliografici
Autori principali: Fang, Tianqing, Zhang, Zhisong, Wang, Xiaoyang, Wang, Rui, Qin, Can, Wan, Yuxuan, Ma, Jun-Yu, Zhang, Ce, Chen, Jiaqi, Li, Xiyun, Wang, Yonglin, Ni, Jingchen, Zheng, Tianshi, Chen, Chun, Yu, Wenhao, Liang, Zhenwen, Zhang, Hongming, Mi, Haitao, Yu, Dong
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2508.00414
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910155148361728
author Fang, Tianqing
Zhang, Zhisong
Wang, Xiaoyang
Wang, Rui
Qin, Can
Wan, Yuxuan
Ma, Jun-Yu
Zhang, Ce
Chen, Jiaqi
Li, Xiyun
Wang, Yonglin
Ni, Jingchen
Zheng, Tianshi
Chen, Chun
Yu, Wenhao
Liang, Zhenwen
Zhang, Hongming
Mi, Haitao
Yu, Dong
author_facet Fang, Tianqing
Zhang, Zhisong
Wang, Xiaoyang
Wang, Rui
Qin, Can
Wan, Yuxuan
Ma, Jun-Yu
Zhang, Ce
Chen, Jiaqi
Li, Xiyun
Wang, Yonglin
Ni, Jingchen
Zheng, Tianshi
Chen, Chun
Yu, Wenhao
Liang, Zhenwen
Zhang, Hongming
Mi, Haitao
Yu, Dong
contents General AI Agents are increasingly recognized as foundational frameworks for the next generation of artificial intelligence, enabling complex reasoning, web interaction, coding, and autonomous research capabilities. However, current agent systems are either closed-source or heavily reliant on a variety of paid APIs and proprietary tools, limiting accessibility and reproducibility for the research community. In this work, we present \textbf{Cognitive Kernel-Pro}, a fully open-source and (to the maximum extent) free multi-module agent framework designed to democratize the development and evaluation of advanced AI agents. Within Cognitive Kernel-Pro, we systematically investigate the curation of high-quality training data for Agent Foundation Models, focusing on the construction of queries, trajectories, and verifiable answers across four key domains: web, file, code, and general reasoning. Furthermore, we explore novel strategies for agent test-time reflection and voting to enhance agent robustness and performance. We evaluate Cognitive Kernel-Pro on GAIA, achieving state-of-the-art results among open-source and free agents. Notably, our 8B-parameter open-source model surpasses previous leading systems such as WebDancer and WebSailor, establishing a new performance standard for accessible, high-capability AI agents. Code is available at https://github.com/Tencent/CognitiveKernel-Pro
format Preprint
id arxiv_https___arxiv_org_abs_2508_00414
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Cognitive Kernel-Pro: A Framework for Deep Research Agents and Agent Foundation Models Training
Fang, Tianqing
Zhang, Zhisong
Wang, Xiaoyang
Wang, Rui
Qin, Can
Wan, Yuxuan
Ma, Jun-Yu
Zhang, Ce
Chen, Jiaqi
Li, Xiyun
Wang, Yonglin
Ni, Jingchen
Zheng, Tianshi
Chen, Chun
Yu, Wenhao
Liang, Zhenwen
Zhang, Hongming
Mi, Haitao
Yu, Dong
Artificial Intelligence
Computation and Language
General AI Agents are increasingly recognized as foundational frameworks for the next generation of artificial intelligence, enabling complex reasoning, web interaction, coding, and autonomous research capabilities. However, current agent systems are either closed-source or heavily reliant on a variety of paid APIs and proprietary tools, limiting accessibility and reproducibility for the research community. In this work, we present \textbf{Cognitive Kernel-Pro}, a fully open-source and (to the maximum extent) free multi-module agent framework designed to democratize the development and evaluation of advanced AI agents. Within Cognitive Kernel-Pro, we systematically investigate the curation of high-quality training data for Agent Foundation Models, focusing on the construction of queries, trajectories, and verifiable answers across four key domains: web, file, code, and general reasoning. Furthermore, we explore novel strategies for agent test-time reflection and voting to enhance agent robustness and performance. We evaluate Cognitive Kernel-Pro on GAIA, achieving state-of-the-art results among open-source and free agents. Notably, our 8B-parameter open-source model surpasses previous leading systems such as WebDancer and WebSailor, establishing a new performance standard for accessible, high-capability AI agents. Code is available at https://github.com/Tencent/CognitiveKernel-Pro
title Cognitive Kernel-Pro: A Framework for Deep Research Agents and Agent Foundation Models Training
topic Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2508.00414