[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-85179-en":3,"doc-seo-85179-105":29,"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":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},85179,962075114765,"Quinn","https://ap-avatar.wpscdn.com/davatar_a8503ba1806abce46bf441b54a3ca4cd",8,"Research & Report","GAE Graph-Augmented Evolution for Scientific Discovery via Reinforcement Optimization","Evolutionary program search guided by large language models (LLMs) has become a key approach for automated scientific discovery, yet existing methods face three limits: structurally blind parent selection, sparse whole-program evaluation rewards, and static mutation operators that cannot adapt during search. GAE (Graph-Augmented Evolution) introduces a graph-structured, reinforcement-optimized framework: typed GNN embeddings, an RL meta-controller for directed parent and mutation choices, and an online GRPO fine-tuning loop to align LLM edits with high-fitness structural improvements. Evaluations on symbolic regression for nonlinear oscillator systems show strong, state-of-the-art out-of-distribution performance.","GAE: Graph-Augmented Evolution for Scientific Discovery via Reinforcement Optimization  \nXuanzhou Chen 1 2 Taoli Cheng 1  \narXiv :2607 . 10 127v 1 [ cs .LG] 11 Jul 2026  \nAbstract  \nEvolutionary program search guided by Large Language Models (LLMs) has emerged as a powerful paradigm for automated scientific discovery. However, current approaches are fundamentally constrained by three bottlenecks: structurally blind parent selection, sparse whole-program evaluation rewards, and static mutation operators that fail to adapt during search. We present GAE (Graph-Augmented Evolution), a framework that resolves these limitations through a tightly coupled, three-pillar architecture. First, a relational graph neural network (GNN) parses programs into typed computation graphs, producing structure-aware embeddings. Second, an RLoptimized meta-controller leverages these embeddings to replace blind evolutionary sampling with a directed policy, dynamically selecting optimal parents and mutation directions based on reward history. Third, an online GRPO finetuning loop continuously updates the LLM mutation operator at test-time using group-normalized evaluation rewards, directly aligning the model’s generation distribution with high-fitness structural edits. We evaluate GAE on a challenging scientific discovery task: symbolic regression for complex nonlinear oscillator systems. By transforming stochastic search into a directed, self-improving trajectory, GAE efficiently discovers closed-form physical equations, consistently matching or outperforming static LLM-driven baselines and achieving state-of-the-art out-ofdistribution performance.  \n1 Shanghai Academy of AI for Science, Shanghai, China 2 School of Electrical and Computer Engineering, Georgia Institute of Technology, Georgia, Unites States. Correspondence to: Taoli Cheng \u003C[chengtaoli.1990@gmail.com](chengtaoli.1990@gmail.com) >.  \nAccepted at ICML 2026 AI for Science Workshop. Copyright 2026 by the author(s) .  \n1. Introduction  \nThe discovery of new algorithms and programs is a central frontier in AI-assisted science. Recent work follows a simple but powerful recipe: combine a large language model with programmatic evaluation and iterative search. FunSearch (Romera-Paredes et al., 2024) first demonstrated that this loop can discover genuinely new mathematical objects, outperforming decades of hand-crafted combinatorial search. AlphaEvolve (Novikov et al., 2025) generalized the principle to complete codebases, achieving state-of-the-art results on hardware scheduling, matrix multiplication, and open problems in math and engineering. Following this line of work, the overall procedure can be viewed as an iterative search framework in which an LLM serves as a program mutation operator that generates candidate code modifications, a task-specific evaluator assigns fitness based on predefined scoring criteria, and a selection mechanism iteratively preserves and refines high-performing programs within the search space.  \nDespite these successes, three interrelated bottlenecks hinder effective evolutionary search, especially in reliably steering program mutations toward high-quality regions of a large and complex program space. (1) Reward sparsity. Each fitness evaluation requires a complete program execution as a full training run in the case of program search, or a numerical integration in symbolic regression. The archive is updated only when a child outperforms the current occupant of its MAP-Elites cell, so the search process receives a single scalar signal indicating whether the candidate improves upon the incumbent in terms of the defined fitness score. Programs that nearly beat an incumbent, for example achieving performance very close to the current best, or that contain partially correct or structurally meaningful sub-expressions, are still discarded if they do not strictly improve the cell. As a result, a large population of evaluated programs is reduced to a sparse sequence of binar","cbCaiqRs3sZphxYZ","https://ap.wps.com/l/cbCaiqRs3sZphxYZ","pdf",4054742,1,12,"English","en",105,"# Abstract\n# Introduction\n## Motivation and Bottlenecks\n## Proposed Contributions","[{\"question\":\"What are the three bottlenecks that current LLM-guided evolutionary search methods face?\",\"answer\":\"They include reward sparsity from binary improvement signals, uninformed parent selection that ignores structural locality, and static LLM mutation operators that do not adapt as the search progresses.\"},{\"question\":\"How does GAE use graph neural networks in its framework?\",\"answer\":\"A relational GNN parses each program into a typed computation graph and produces structure-aware embeddings that support later decision-making in the search process.\"},{\"question\":\"How does GAE improve mutation quality during the search?\",\"answer\":\"An RL-optimized meta-controller directs parent and mutation choices based on reward history, and an online GRPO fine-tuning loop updates the LLM mutation operator at test time using group-normalized evaluation rewards to better match high-fitness structural edits.\"}]",1784201574,30,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"gae-graph-augmented-evolution-for-scientific-discovery-via-reinforcement-optimization","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/gae-graph-augmented-evolution-for-scientific-discovery-via-reinforcement-optimization/85179/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","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 are the three bottlenecks that current LLM-guided evolutionary search methods face?","Question",{"text":75,"@type":76},"They include reward sparsity from binary improvement signals, uninformed parent selection that ignores structural locality, and static LLM mutation operators that do not adapt as the search progresses.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does GAE use graph neural networks in its framework?",{"text":80,"@type":76},"A relational GNN parses each program into a typed computation graph and produces structure-aware embeddings that support later decision-making in the search process.",{"name":82,"@type":73,"acceptedAnswer":83},"How does GAE improve mutation quality during the search?",{"text":84,"@type":76},"An RL-optimized meta-controller directs parent and mutation choices based on reward history, and an online GRPO fine-tuning loop updates the LLM mutation operator at test time using group-normalized evaluation rewards to better match high-fitness structural edits.","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":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":28,"slug":121},"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]