Infant Agent: A Tool-Integrated, Logic-Driven Agent with Cost-Effective API Usage

Despite the impressive capabilities of large language models (LLMs), they currently exhibit two primary limitations, \textbf{\uppercase\expandafter{\romannumeral 1}}: They struggle to \textbf{autonomously solve the real world engineering problem}. \textbf{\uppercase\expandafter{\romannumeral 2}}: Th...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:arXiv.org 2024-11
Hauptverfasser: Lei, Bin, Li, Yuchen, Zeng, Yiming, Ren, Tao, Luo, Yi, Shi, Tianyu, Gao, Zitian, Hu, Zeyu, Kang, Weitai, Chen, Qiuwu
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title arXiv.org
container_volume
creator Lei, Bin
Li, Yuchen
Zeng, Yiming
Ren, Tao
Luo, Yi
Shi, Tianyu
Gao, Zitian
Hu, Zeyu
Kang, Weitai
Chen, Qiuwu
description Despite the impressive capabilities of large language models (LLMs), they currently exhibit two primary limitations, \textbf{\uppercase\expandafter{\romannumeral 1}}: They struggle to \textbf{autonomously solve the real world engineering problem}. \textbf{\uppercase\expandafter{\romannumeral 2}}: They remain \textbf{challenged in reasoning through complex logic problems}. To address these challenges, we developed the \textsc{Infant Agent}, integrating task-aware functions, operators, a hierarchical management system, and a memory retrieval mechanism. Together, these components enable large language models to sustain extended reasoning processes and handle complex, multi-step tasks efficiently, all while significantly reducing API costs. Using the \textsc{Infant Agent}, GPT-4o's accuracy on the SWE-bench-lite dataset rises from \(\mathbf{0.33\%}\) to \(\mathbf{30\%}\), and in the AIME-2024 mathematics competition, it increases GPT-4o's accuracy from \(\mathbf{13.3\%}\) to \(\mathbf{37\%}\).
format Article
fullrecord <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_3124192280</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>3124192280</sourcerecordid><originalsourceid>FETCH-proquest_journals_31241922803</originalsourceid><addsrcrecordid>eNqNysEKgjAcgPERBEn5DoOuDeamZd3EjIQOBXaWYf8tRbbaZr1-Qj1Ap-_w-yYoYJxHJI0Zm6HQuY5SytYbliQ8QJdSS6E9zhRov8MZrozpSak9KCs83Fb4ZFTbkL1tX6C_G363_o5z4zwppITGj4Szc4mvTihYoKkUvYPw1zlaHooqP5KHNc8BnK87M1g9Us0jFkdbxlLK_7s-oz09Ug</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>3124192280</pqid></control><display><type>article</type><title>Infant Agent: A Tool-Integrated, Logic-Driven Agent with Cost-Effective API Usage</title><source>Free E- Journals</source><creator>Lei, Bin ; Li, Yuchen ; Zeng, Yiming ; Ren, Tao ; Luo, Yi ; Shi, Tianyu ; Gao, Zitian ; Hu, Zeyu ; Kang, Weitai ; Chen, Qiuwu</creator><creatorcontrib>Lei, Bin ; Li, Yuchen ; Zeng, Yiming ; Ren, Tao ; Luo, Yi ; Shi, Tianyu ; Gao, Zitian ; Hu, Zeyu ; Kang, Weitai ; Chen, Qiuwu</creatorcontrib><description>Despite the impressive capabilities of large language models (LLMs), they currently exhibit two primary limitations, \textbf{\uppercase\expandafter{\romannumeral 1}}: They struggle to \textbf{autonomously solve the real world engineering problem}. \textbf{\uppercase\expandafter{\romannumeral 2}}: They remain \textbf{challenged in reasoning through complex logic problems}. To address these challenges, we developed the \textsc{Infant Agent}, integrating task-aware functions, operators, a hierarchical management system, and a memory retrieval mechanism. Together, these components enable large language models to sustain extended reasoning processes and handle complex, multi-step tasks efficiently, all while significantly reducing API costs. Using the \textsc{Infant Agent}, GPT-4o's accuracy on the SWE-bench-lite dataset rises from \(\mathbf{0.33\%}\) to \(\mathbf{30\%}\), and in the AIME-2024 mathematics competition, it increases GPT-4o's accuracy from \(\mathbf{13.3\%}\) to \(\mathbf{37\%}\).</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Accuracy ; Infants ; Large language models ; Operators (mathematics) ; Reasoning ; Task complexity</subject><ispartof>arXiv.org, 2024-11</ispartof><rights>2024. This work is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>780,784</link.rule.ids></links><search><creatorcontrib>Lei, Bin</creatorcontrib><creatorcontrib>Li, Yuchen</creatorcontrib><creatorcontrib>Zeng, Yiming</creatorcontrib><creatorcontrib>Ren, Tao</creatorcontrib><creatorcontrib>Luo, Yi</creatorcontrib><creatorcontrib>Shi, Tianyu</creatorcontrib><creatorcontrib>Gao, Zitian</creatorcontrib><creatorcontrib>Hu, Zeyu</creatorcontrib><creatorcontrib>Kang, Weitai</creatorcontrib><creatorcontrib>Chen, Qiuwu</creatorcontrib><title>Infant Agent: A Tool-Integrated, Logic-Driven Agent with Cost-Effective API Usage</title><title>arXiv.org</title><description>Despite the impressive capabilities of large language models (LLMs), they currently exhibit two primary limitations, \textbf{\uppercase\expandafter{\romannumeral 1}}: They struggle to \textbf{autonomously solve the real world engineering problem}. \textbf{\uppercase\expandafter{\romannumeral 2}}: They remain \textbf{challenged in reasoning through complex logic problems}. To address these challenges, we developed the \textsc{Infant Agent}, integrating task-aware functions, operators, a hierarchical management system, and a memory retrieval mechanism. Together, these components enable large language models to sustain extended reasoning processes and handle complex, multi-step tasks efficiently, all while significantly reducing API costs. Using the \textsc{Infant Agent}, GPT-4o's accuracy on the SWE-bench-lite dataset rises from \(\mathbf{0.33\%}\) to \(\mathbf{30\%}\), and in the AIME-2024 mathematics competition, it increases GPT-4o's accuracy from \(\mathbf{13.3\%}\) to \(\mathbf{37\%}\).</description><subject>Accuracy</subject><subject>Infants</subject><subject>Large language models</subject><subject>Operators (mathematics)</subject><subject>Reasoning</subject><subject>Task complexity</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>ABUWG</sourceid><sourceid>AFKRA</sourceid><sourceid>AZQEC</sourceid><sourceid>BENPR</sourceid><sourceid>CCPQU</sourceid><sourceid>DWQXO</sourceid><recordid>eNqNysEKgjAcgPERBEn5DoOuDeamZd3EjIQOBXaWYf8tRbbaZr1-Qj1Ap-_w-yYoYJxHJI0Zm6HQuY5SytYbliQ8QJdSS6E9zhRov8MZrozpSak9KCs83Fb4ZFTbkL1tX6C_G363_o5z4zwppITGj4Szc4mvTihYoKkUvYPw1zlaHooqP5KHNc8BnK87M1g9Us0jFkdbxlLK_7s-oz09Ug</recordid><startdate>20241102</startdate><enddate>20241102</enddate><creator>Lei, Bin</creator><creator>Li, Yuchen</creator><creator>Zeng, Yiming</creator><creator>Ren, Tao</creator><creator>Luo, Yi</creator><creator>Shi, Tianyu</creator><creator>Gao, Zitian</creator><creator>Hu, Zeyu</creator><creator>Kang, Weitai</creator><creator>Chen, Qiuwu</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope></search><sort><creationdate>20241102</creationdate><title>Infant Agent: A Tool-Integrated, Logic-Driven Agent with Cost-Effective API Usage</title><author>Lei, Bin ; Li, Yuchen ; Zeng, Yiming ; Ren, Tao ; Luo, Yi ; Shi, Tianyu ; Gao, Zitian ; Hu, Zeyu ; Kang, Weitai ; Chen, Qiuwu</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_31241922803</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Accuracy</topic><topic>Infants</topic><topic>Large language models</topic><topic>Operators (mathematics)</topic><topic>Reasoning</topic><topic>Task complexity</topic><toplevel>online_resources</toplevel><creatorcontrib>Lei, Bin</creatorcontrib><creatorcontrib>Li, Yuchen</creatorcontrib><creatorcontrib>Zeng, Yiming</creatorcontrib><creatorcontrib>Ren, Tao</creatorcontrib><creatorcontrib>Luo, Yi</creatorcontrib><creatorcontrib>Shi, Tianyu</creatorcontrib><creatorcontrib>Gao, Zitian</creatorcontrib><creatorcontrib>Hu, Zeyu</creatorcontrib><creatorcontrib>Kang, Weitai</creatorcontrib><creatorcontrib>Chen, Qiuwu</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science &amp; Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Lei, Bin</au><au>Li, Yuchen</au><au>Zeng, Yiming</au><au>Ren, Tao</au><au>Luo, Yi</au><au>Shi, Tianyu</au><au>Gao, Zitian</au><au>Hu, Zeyu</au><au>Kang, Weitai</au><au>Chen, Qiuwu</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>Infant Agent: A Tool-Integrated, Logic-Driven Agent with Cost-Effective API Usage</atitle><jtitle>arXiv.org</jtitle><date>2024-11-02</date><risdate>2024</risdate><eissn>2331-8422</eissn><abstract>Despite the impressive capabilities of large language models (LLMs), they currently exhibit two primary limitations, \textbf{\uppercase\expandafter{\romannumeral 1}}: They struggle to \textbf{autonomously solve the real world engineering problem}. \textbf{\uppercase\expandafter{\romannumeral 2}}: They remain \textbf{challenged in reasoning through complex logic problems}. To address these challenges, we developed the \textsc{Infant Agent}, integrating task-aware functions, operators, a hierarchical management system, and a memory retrieval mechanism. Together, these components enable large language models to sustain extended reasoning processes and handle complex, multi-step tasks efficiently, all while significantly reducing API costs. Using the \textsc{Infant Agent}, GPT-4o's accuracy on the SWE-bench-lite dataset rises from \(\mathbf{0.33\%}\) to \(\mathbf{30\%}\), and in the AIME-2024 mathematics competition, it increases GPT-4o's accuracy from \(\mathbf{13.3\%}\) to \(\mathbf{37\%}\).</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier EISSN: 2331-8422
ispartof arXiv.org, 2024-11
issn 2331-8422
language eng
recordid cdi_proquest_journals_3124192280
source Free E- Journals
subjects Accuracy
Infants
Large language models
Operators (mathematics)
Reasoning
Task complexity
title Infant Agent: A Tool-Integrated, Logic-Driven Agent with Cost-Effective API Usage
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-06T16%3A32%3A13IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=document&rft.atitle=Infant%20Agent:%20A%20Tool-Integrated,%20Logic-Driven%20Agent%20with%20Cost-Effective%20API%20Usage&rft.jtitle=arXiv.org&rft.au=Lei,%20Bin&rft.date=2024-11-02&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E3124192280%3C/proquest%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=3124192280&rft_id=info:pmid/&rfr_iscdi=true