Deprecated: The each() function is deprecated. This message will be suppressed on further calls in /home/zhenxiangba/zhenxiangba.com/public_html/phproxy-improved-master/index.php on line 456
Paper page - GATE: Graph-based Adaptive Tool Evolution Across Diverse Tasks
[go: Go Back, main page]

https://github.com/ayanami2003/GATE.

\n","updatedAt":"2025-04-28T07:05:12.242Z","author":{"_id":"63fb6e281b4b1bd4e7ffc5be","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/63fb6e281b4b1bd4e7ffc5be/aiRu_bulgnxvEMrjipGoQ.jpeg","fullname":"Xiao Liu","name":"lx865712528","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":13,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.8688520789146423},"editors":["lx865712528"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/63fb6e281b4b1bd4e7ffc5be/aiRu_bulgnxvEMrjipGoQ.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2502.14848","authors":[{"_id":"680f286d8e61f53b0fd8116c","user":{"_id":"66adf5cc0c6056d9f4dc308f","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/66adf5cc0c6056d9f4dc308f/mVKo06P7M1qf6RYNG-c2i.jpeg","isPro":false,"fullname":"Jane Luo","user":"Luo2003","type":"user"},"name":"Jianwen Luo","status":"claimed_verified","statusLastChangedAt":"2025-09-22T02:53:02.659Z","hidden":false},{"_id":"680f286d8e61f53b0fd8116d","user":{"_id":"6514599ee31c0e2e3dfb5c9c","avatarUrl":"/avatars/3c3ebd14d228c4c439da542cf8ff20a8.svg","isPro":false,"fullname":"ymh233","user":"ymh233","type":"user"},"name":"Yiming Huang","status":"claimed_verified","statusLastChangedAt":"2025-09-23T02:43:12.874Z","hidden":false},{"_id":"680f286d8e61f53b0fd8116e","name":"Jinxiang Meng","hidden":false},{"_id":"680f286d8e61f53b0fd8116f","name":"Fangyu Lei","hidden":false},{"_id":"680f286d8e61f53b0fd81170","name":"Shizhu He","hidden":false},{"_id":"680f286d8e61f53b0fd81171","user":{"_id":"63fb6e281b4b1bd4e7ffc5be","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/63fb6e281b4b1bd4e7ffc5be/aiRu_bulgnxvEMrjipGoQ.jpeg","isPro":false,"fullname":"Xiao Liu","user":"lx865712528","type":"user"},"name":"Xiao Liu","status":"claimed_verified","statusLastChangedAt":"2025-04-28T07:38:09.184Z","hidden":false},{"_id":"680f286d8e61f53b0fd81172","name":"Shanshan Jiang","hidden":false},{"_id":"680f286d8e61f53b0fd81173","name":"Bin Dong","hidden":false},{"_id":"680f286d8e61f53b0fd81174","name":"Jun Zhao","hidden":false},{"_id":"680f286d8e61f53b0fd81175","name":"Kang Liu","hidden":false}],"publishedAt":"2025-02-20T18:56:03.000Z","title":"GATE: Graph-based Adaptive Tool Evolution Across Diverse Tasks","summary":"Large Language Models (LLMs) have shown great promise in tool-making, yet\nexisting frameworks often struggle to efficiently construct reliable toolsets\nand are limited to single-task settings. To address these challenges, we\npropose GATE (Graph-based Adaptive Tool Evolution), an adaptive framework that\ndynamically constructs and evolves a hierarchical graph of reusable tools\nacross multiple scenarios. We evaluate GATE on open-ended tasks (Minecraft),\nagent-based tasks (TextCraft, DABench), and code generation tasks (MATH, Date,\nTabMWP). Our results show that GATE achieves up to 4.3x faster milestone\ncompletion in Minecraft compared to the previous SOTA, and provides an average\nimprovement of 9.23% over existing tool-making methods in code generation tasks\nand 10.03% in agent tasks. GATE demonstrates the power of adaptive evolution,\nbalancing tool quantity, complexity, and functionality while maintaining high\nefficiency. Code and data are available at\nhttps://github.com/ayanami2003/GATE.","upvotes":1,"discussionId":"680f28778e61f53b0fd813e4","githubRepo":"https://github.com/ayanami2003/gate","githubRepoAddedBy":"auto","ai_summary":"GATE, a graph-based adaptive framework, dynamically evolves toolsets across various tasks, significantly enhancing efficiency and performance compared to existing methods.","ai_keywords":["Large Language Models","LLMs","tool-making","adaptive framework","hierarchical graph","reusable tools","open-ended tasks","Minecraft","agent-based tasks","TextCraft","DABench","code generation tasks","MATH","Date","TabMWP","adaptive evolution","milestone completion","tool quantity","tool complexity","tool functionality"],"githubStars":4},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"63fb6e281b4b1bd4e7ffc5be","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/63fb6e281b4b1bd4e7ffc5be/aiRu_bulgnxvEMrjipGoQ.jpeg","isPro":false,"fullname":"Xiao Liu","user":"lx865712528","type":"user"}],"acceptLanguages":["*"]}">
Papers
arxiv:2502.14848

GATE: Graph-based Adaptive Tool Evolution Across Diverse Tasks

Published on Feb 20, 2025
Authors:
,
,
,
,
,
,

Abstract

GATE, a graph-based adaptive framework, dynamically evolves toolsets across various tasks, significantly enhancing efficiency and performance compared to existing methods.

AI-generated summary

Large Language Models (LLMs) have shown great promise in tool-making, yet existing frameworks often struggle to efficiently construct reliable toolsets and are limited to single-task settings. To address these challenges, we propose GATE (Graph-based Adaptive Tool Evolution), an adaptive framework that dynamically constructs and evolves a hierarchical graph of reusable tools across multiple scenarios. We evaluate GATE on open-ended tasks (Minecraft), agent-based tasks (TextCraft, DABench), and code generation tasks (MATH, Date, TabMWP). Our results show that GATE achieves up to 4.3x faster milestone completion in Minecraft compared to the previous SOTA, and provides an average improvement of 9.23% over existing tool-making methods in code generation tasks and 10.03% in agent tasks. GATE demonstrates the power of adaptive evolution, balancing tool quantity, complexity, and functionality while maintaining high efficiency. Code and data are available at https://github.com/ayanami2003/GATE.

Community

Paper author

Large Language Models (LLMs) have shown great promise in tool-making, yet existing frameworks often struggle to efficiently construct reliable toolsets and are limited to single-task settings. To address these challenges, we propose GATE (Graph-based Adaptive Tool Evolution), an adaptive framework that dynamically constructs and evolves a hierarchical graph of reusable tools across multiple scenarios. We evaluate GATE on open-ended tasks (Minecraft), agent-based tasks (TextCraft, DABench), and code generation tasks (MATH, Date, TabMWP). Our results show that GATE achieves up to 4.3x faster milestone completion in Minecraft compared to the previous SOTA, and provides an average improvement of 9.23% over existing tool-making methods in code generation tasks and 10.03% in agent tasks. GATE demonstrates the power of adaptive evolution, balancing tool quantity, complexity, and functionality while maintaining high efficiency. Code and data are available at https://github.com/ayanami2003/GATE.

Sign up or log in to comment

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2502.14848 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2502.14848 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2502.14848 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.