[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-85895-en":3,"doc-seo-85895-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":13,"seo_description":14,"update_tm":28,"read_time":29},85895,687197207639,"Asher","https://ap-avatar.wpscdn.com/davatar_a8503ba1806abce46bf441b54a3ca4cd",8,"Research & Report","Co4ICF Co-evolving Physics-Informed Surrogate and RL-based Pulse Optimizer for Inertial Confinement Fusion","Offline-trained surrogates for Inertial Confinement Fusion (ICF) can fail when iterative optimizers drive designs into out-of-distribution (OOD) regions, making surrogate predictions unreliable. Co4ICF introduces a co-evolving framework that couples a physics-informed surrogate with a PPO-based pulse optimizer. The surrogate is periodically fine-tuned on optimizer-induced trajectories, correcting extrapolation errors as the input distribution shifts. In 1D-MULTI, it reaches 146.1% normalized yield and achieves 246.9% in 2D-MULTI cross-fidelity evaluation without 2D training, supported by budget-matched ablations. A large-scale MULTI-IFE dataset is released for benchmarking.","Co4ICF: Co-evolving Physics-Informed Surrogate and RL-based Pulse Optimizer for Inertial Confinement Fusion  \nJiatong Zhao∗  \n[zhaojiatong@sjtu.edu.cn](zhaojiatong@sjtu.edu.cn)[ ](zhaojiatong@sjtu.edu.cn)School of Physics and Astronomy Zhiyuan College Shanghai Jiao Tong University Shanghai, China  \nTengyue Zhang∗  \n[zhangty_23@sjtu.edu.cn](zhangty_23@sjtu.edu.cn)[ ](zhangty_23@sjtu.edu.cn)School of Physics and Astronomy Zhiyuan College Shanghai Jiao Tong University Shanghai, China  \nYuhan Wang∗  \n[wang.yuhan@sjtu.edu.cn](wang.yuhan@sjtu.edu.cn)[ ](wang.yuhan@sjtu.edu.cn)School of Physics and Astronomy Zhiyuan College Shanghai Jiao Tong University Shanghai, China  \nFuyuan Wu  \n[fuyuan.wu@sjtu.edu.cn](fuyuan.wu@sjtu.edu.cn)[ ](fuyuan.wu@sjtu.edu.cn)School of Physics and Astronomy Shanghai Jiao Tong University Shanghai, China  \nJunchi Yan†  \n[yanjunchi@sjtu.edu.cn](yanjunchi@sjtu.edu.cn)[ ](yanjunchi@sjtu.edu.cn)School of Artificial Intelligence Shanghai Jiao Tong University Shanghai, China  \narXiv :2607 . 10366v 1 [ cs .AI] 11 Jul 2026  \nFigure 1: Performance of Co4ICF in laser pulse optimization. (a) Fusion Yield: The Co4ICF framework (blue solid line) dynamically calibrates the surrogate, reaching 146.1% normalized yield in the 1D-MULTI optimization loop, i.e., a 46.1% improvement over the designed baseline (red dotted line). This outperforms the static surrogate approach (orange dashed line, +15.4%), which suffers from distribution shift. (Re-evaluated by 1D-MULTI; see Table 3 for final direct 2D-MULTI evaluation.) (b) Pulse Optimization: Comparison between the initial random pulse samples (top rows) and the final RL-optimized waveforms (bottom rows), showing the pulse structures found by the policy.  \nAbstract  \nOffline-trained surrogates for Inertial Confinement Fusion (ICF) suffer a well-known failure mode that iterative optimizers drive inputs into out-of-distribution (OOD) regions where predictions become unreliable. Here we present Co4ICF, a co-evolving framework that couples a physics-informed surrogate with a PPO-based  \n∗ These authors contributed equally to this research.†Corresponding author.  \nThis work is licensed under a Creative Commons Attribution 4 .0 International License. KDD’26, Jeju Island, Republic of Korea  \n© 2026 Copyright held by the owner/author(s) .  \nACM ISBN 979-8-4007-2259-2/2026/08  \n[https://doi.org/10.1145/3770855.3818998](https://doi.org/10.1145/3770855.3818998)  \npulse optimizer. The surrogate is iteratively fine-tuned on policyinduced trajectories, correcting extrapolation errors as the optimizer shifts the input distribution; the optimizer queries this evolving surrogate as a fast environment. In the 1D MULTI environment, Co4ICF achieves 146.1% normalized yield based on current laser design baseline; as a post-hoc cross-fidelity check, the optimized pulse further attains 246.9% normalized yield when directly evaluated in 2D-MULTI without any 2D training or fine-tuning. Budget-matched ablations support that the gains are not explained solely by additional simulation data and are consistent with the coevolving mechanism playing a key role. We release a large-scale MULTI-IFE simulation dataset to support future benchmarking.  \nKDD’26, August 09–13, 2026, Jeju Island, Republic of Korea Jiatong Zhao, Tengyue Zhang, Yuhan Wang, Fuyuan Wu and Junchi Yan  \nCCS Concepts  \n• Applied computing → Physics; • Computing methodologies → Modeling methodologies; Machine learning; Artificial intelligence.  \nKeywords  \nInertial Confinement Fusion, Surrogate Modeling, Reinforcement Learning, Pulse Optimization  \nACM Reference Format:  \nJiatong Zhao, Tengyue Zhang, Yuhan Wang, Fuyuan Wu, and Junchi Yan.  \n2026. Co4ICF: Co-evolving Physics-Informed Surrogate andRL-based Pulse Optimizer for Inertial Confinement Fusion. In Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.2 (KDD’26), August 09–13, 2026, Jeju Island, Republic ofKorea. ACM, New York, NY, USA, 11 pages. [https://","cbCaiiwV9xPFp40r","https://ap.wps.com/l/cbCaiiwV9xPFp40r","pdf",5798151,3,1,11,"English","en",105,"# Introduction\n## Problem: OOD failure and cost of high-fidelity simulation\n## Proposed solution: Co4ICF co-evolving loop\n## Contributions and evaluation setup","[{\"question\":\"What problem does Co4ICF address in ICF surrogate-based optimization?\",\"answer\":\"It addresses the failure mode of offline-trained surrogates when iterative optimizers push inputs into out-of-distribution (OOD) regions, causing unreliable predictions and also concerns about physical consistency.\"},{\"question\":\"How does the Co4ICF framework combine the surrogate and the RL pulse optimizer?\",\"answer\":\"Co4ICF couples a physics-informed surrogate (replacing expensive MULTI simulations) with a PPO-based pulse designer. The optimizer queries the surrogate as a fast environment, while the surrogate is periodically fine-tuned on policy-induced trajectories.\"},{\"question\":\"What performance results does Co4ICF report across 1D and 2D evaluations?\",\"answer\":\"In 1D-MULTI, Co4ICF reaches 146.1% normalized yield, improving over a designed baseline. A post-hoc cross-fidelity check shows the optimized pulse attains 246.9% normalized yield when directly evaluated in 2D-MULTI without any 2D training or fine-tuning.\"}]",1784207010,28,{"code":4,"msg":31,"data":32},"ok",{"site_id":25,"language":24,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"co4icf-co-evolving-physics-informed-surrogate-and-rl-based-pulse-optimizer-for-inertial-confinement-fusion","",{"@graph":36,"@context":85},[37,53,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,50],{"item":41,"name":42,"@type":43,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":20},"https://docshare.wps.com/document/research-report/",{"item":51,"name":13,"@type":43,"position":52},"https://docshare.wps.com/document/co4icf-co-evolving-physics-informed-surrogate-and-rl-based-pulse-optimizer-for-inertial-confinement-fusion/85895/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":24,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":41,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-25","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does Co4ICF address in ICF surrogate-based optimization?","Question",{"text":75,"@type":76},"It addresses the failure mode of offline-trained surrogates when iterative optimizers push inputs into out-of-distribution (OOD) regions, causing unreliable predictions and also concerns about physical consistency.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the Co4ICF framework combine the surrogate and the RL pulse optimizer?",{"text":80,"@type":76},"Co4ICF couples a physics-informed surrogate (replacing expensive MULTI simulations) with a PPO-based pulse designer. The optimizer queries the surrogate as a fast environment, while the surrogate is periodically fine-tuned on policy-induced trajectories.",{"name":82,"@type":73,"acceptedAnswer":83},"What performance results does Co4ICF report across 1D and 2D evaluations?",{"text":84,"@type":76},"In 1D-MULTI, Co4ICF reaches 146.1% normalized yield, improving over a designed baseline. A post-hoc cross-fidelity check shows the optimized pulse attains 246.9% normalized yield when directly evaluated in 2D-MULTI without any 2D training or fine-tuning.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]