Intern-S1: Un Modello Fondamentale Scientifico Multimodale

Intern-S1: A Scientific Multimodal Foundation Model

August 21, 2025

Autori: Lei Bai, Zhongrui Cai, Maosong Cao, Weihan Cao, Chiyu Chen, Haojiong Chen, Kai Chen, Pengcheng Chen, Ying Chen, Yongkang Chen, Yu Cheng, Yu Cheng, Pei Chu, Tao Chu, Erfei Cui, Ganqu Cui, Long Cui, Ziyun Cui, Nianchen Deng, Ning Ding, Nanqin Dong, Peijie Dong, Shihan Dou, Sinan Du, Haodong Duan, Caihua Fan, Ben Gao, Changjiang Gao, Jianfei Gao, Songyang Gao, Yang Gao, Zhangwei Gao, Jiaye Ge, Qiming Ge, Lixin Gu, Yuzhe Gu, Aijia Guo, Qipeng Guo, Xu Guo, Conghui He, Junjun He, Yili Hong, Siyuan Hou, Caiyu Hu, Hanglei Hu, Jucheng Hu, Ming Hu, Zhouqi Hua, Haian Huang, Junhao Huang, Xu Huang, Zixian Huang, Zhe Jiang, Lingkai Kong, Linyang Li, Peiji Li, Pengze Li, Shuaibin Li, Tianbin Li, Wei Li, Yuqiang Li, Dahua Lin, Junyao Lin, Tianyi Lin, Zhishan Lin, Hongwei Liu, Jiangning Liu, Jiyao Liu, Junnan Liu, Kai Liu, Kaiwen Liu, Kuikun Liu, Shichun Liu, Shudong Liu, Wei Liu, Xinyao Liu, Yuhong Liu, Zhan Liu, Yinquan Lu, Haijun Lv, Hongxia Lv, Huijie Lv, Qidang Lv, Ying Lv, Chengqi Lyu, Chenglong Ma, Jianpeng Ma, Ren Ma, Runmin Ma, Runyuan Ma, Xinzhu Ma, Yichuan Ma, Zihan Ma, Sixuan Mi, Junzhi Ning, Wenchang Ning, Xinle Pang, Jiahui Peng, Runyu Peng, Yu Qiao, Jiantao Qiu, Xiaoye Qu, Yuan Qu, Yuchen Ren, Fukai Shang, Wenqi Shao, Junhao Shen, Shuaike Shen, Chunfeng Song, Demin Song, Diping Song, Chenlin Su, Weijie Su, Weigao Sun, Yu Sun, Qian Tan, Cheng Tang, Huanze Tang, Kexian Tang, Shixiang Tang, Jian Tong, Aoran Wang, Bin Wang, Dong Wang, Lintao Wang, Rui Wang, Weiyun Wang, Wenhai Wang, Yi Wang, Ziyi Wang, Ling-I Wu, Wen Wu, Yue Wu, Zijian Wu, Linchen Xiao, Shuhao Xing, Chao Xu, Huihui Xu, Jun Xu, Ruiliang Xu, Wanghan Xu, GanLin Yang, Yuming Yang, Haochen Ye, Jin Ye, Shenglong Ye, Jia Yu, Jiashuo Yu, Jing Yu, Fei Yuan, Bo Zhang, Chao Zhang, Chen Zhang, Hongjie Zhang, Jin Zhang, Qiaosheng Zhang, Qiuyinzhe Zhang, Songyang Zhang, Taolin Zhang, Wenlong Zhang, Wenwei Zhang, Yechen Zhang, Ziyang Zhang, Haiteng Zhao, Qian Zhao, Xiangyu Zhao, Xiangyu Zhao, Bowen Zhou, Dongzhan Zhou, Peiheng Zhou, Yuhao Zhou, Yunhua Zhou, Dongsheng Zhu, Lin Zhu, Yicheng Zou

cs.AI

Abstract

Negli ultimi anni, una moltitudine di modelli di base open-source è emersa, ottenendo progressi significativi in alcuni campi ampiamente seguiti, con prestazioni molto vicine a quelle dei modelli closed-source. Tuttavia, in campi scientifici professionali ad alto valore ma più impegnativi, si continua a fare affidamento su modelli specializzati o i progressi dei modelli di base generali rimangono significativamente indietro rispetto a quelli nelle aree più popolari, risultando ben lontani dall’essere sufficienti per trasformare la ricerca scientifica e lasciando un divario sostanziale tra i modelli open-source e quelli closed-source in questi ambiti scientifici. Per ridurre questo divario e fare un ulteriore passo verso l’Intelligenza Artificiale Generale (AGI), presentiamo Intern-S1, uno specialista generalista dotato di capacità di comprensione e ragionamento generale con competenze per analizzare dati multimodali scientifici. Intern-S1 è un modello multimodale Mixture-of-Experts (MoE) con 28 miliardi di parametri attivati e 241 miliardi di parametri totali, pre-addestrato continuamente su 5T di token, inclusi oltre 2.5T di token provenienti da domini scientifici. Nella fase di post-addestramento, Intern-S1 viene sottoposto a un apprendimento per rinforzo (RL) offline e poi online in InternBootCamp, dove proponiamo il Mixture-of-Rewards (MoR) per sinergizzare l’addestramento RL su più di 1000 task simultaneamente. Attraverso innovazioni integrate negli algoritmi, nei dati e nei sistemi di addestramento, Intern-S1 ha raggiunto prestazioni di primo livello nell’addestramento RL online. Su benchmark di valutazione completi, Intern-S1 dimostra prestazioni competitive nei task di ragionamento generale tra i modelli open-source e supera significativamente i modelli open-source nei domini scientifici, superando i modelli closed-source all’avanguardia in task professionali come la pianificazione della sintesi molecolare, la previsione delle condizioni di reazione e la previsione delle stabilità termodinamiche dei cristalli. I nostri modelli sono disponibili su https://huggingface.co/internlm/Intern-S1.

English

In recent years, a plethora of open-source foundation models have emerged, achieving remarkable progress in some widely attended fields, with performance being quite close to that of closed-source models. However, in high-value but more challenging scientific professional fields, either the fields still rely on expert models, or the progress of general foundation models lags significantly compared to those in popular areas, far from sufficient for transforming scientific research and leaving substantial gap between open-source models and closed-source models in these scientific domains. To mitigate this gap and explore a step further toward Artificial General Intelligence (AGI), we introduce Intern-S1, a specialized generalist equipped with general understanding and reasoning capabilities with expertise to analyze multiple science modal data. Intern-S1 is a multimodal Mixture-of-Experts (MoE) model with 28 billion activated parameters and 241 billion total parameters, continually pre-trained on 5T tokens, including over 2.5T tokens from scientific domains. In the post-training stage, Intern-S1 undergoes offline and then online reinforcement learning (RL) in InternBootCamp, where we propose Mixture-of-Rewards (MoR) to synergize the RL training on more than 1000 tasks simultaneously. Through integrated innovations in algorithms, data, and training systems, Intern-S1 achieved top-tier performance in online RL training.On comprehensive evaluation benchmarks, Intern-S1 demonstrates competitive performance on general reasoning tasks among open-source models and significantly outperforms open-source models in scientific domains, surpassing closed-source state-of-the-art models in professional tasks, such as molecular synthesis planning, reaction condition prediction, predicting thermodynamic stabilities for crystals. Our models are available at https://huggingface.co/internlm/Intern-S1.