EcoAssistant: Using LLM Assistant More Affordably and Accurately

Today, users ask Large language models (LLMs) as assistants to answer queries that require external knowledge; they ask about the weather in a specific city, about stock prices, and even about where specific locations are within their neighborhood. These queries require the LLM to produce code that...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Zhang, Jieyu, Krishna, Ranjay, Awadallah, Ahmed H, Wang, Chi
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Zhang, Jieyu
Krishna, Ranjay
Awadallah, Ahmed H
Wang, Chi
description Today, users ask Large language models (LLMs) as assistants to answer queries that require external knowledge; they ask about the weather in a specific city, about stock prices, and even about where specific locations are within their neighborhood. These queries require the LLM to produce code that invokes external APIs to answer the user's question, yet LLMs rarely produce correct code on the first try, requiring iterative code refinement upon execution results. In addition, using LLM assistants to support high query volumes can be expensive. In this work, we contribute a framework, EcoAssistant, that enables LLMs to answer code-driven queries more affordably and accurately. EcoAssistant contains three components. First, it allows the LLM assistants to converse with an automatic code executor to iteratively refine code or to produce answers based on the execution results. Second, we use a hierarchy of LLM assistants, which attempts to answer the query with weaker, cheaper LLMs before backing off to stronger, expensive ones. Third, we retrieve solutions from past successful queries as in-context demonstrations to help subsequent queries. Empirically, we show that EcoAssistant offers distinct advantages for affordability and accuracy, surpassing GPT-4 by 10 points of success rate with less than 50% of GPT-4's cost.
doi_str_mv 10.48550/arxiv.2310.03046
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2310_03046</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2310_03046</sourcerecordid><originalsourceid>FETCH-LOGICAL-a676-d116d1d10c4798dc79fc42622d1ddc709d587fd5b4adbed92fe9cfe63ef8fcd83</originalsourceid><addsrcrecordid>eNo9j81KxDAURrNxIaMP4Mq8QMf8NU1cWYbxBzq4GdflNjdXCrWVpIp9e-sorj6-szhwGLuSYmtcWYobSF_951bpFQgtjD1nd_sw1Tn3eYZxvuUvuR9fedMc-D_khylFXhNNCaEbFg4j8jqEjwRzHJYLdkYw5Hj5txt2vN8fd49F8_zwtKubAmxlC5TSokQpgqm8w1B5CkZZpVa4PuGxdBVh2RnALqJXFH2gaHUkRwGd3rDrX-0poX1P_Rukpf1JaU8p-hvggER6</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>EcoAssistant: Using LLM Assistant More Affordably and Accurately</title><source>arXiv.org</source><creator>Zhang, Jieyu ; Krishna, Ranjay ; Awadallah, Ahmed H ; Wang, Chi</creator><creatorcontrib>Zhang, Jieyu ; Krishna, Ranjay ; Awadallah, Ahmed H ; Wang, Chi</creatorcontrib><description>Today, users ask Large language models (LLMs) as assistants to answer queries that require external knowledge; they ask about the weather in a specific city, about stock prices, and even about where specific locations are within their neighborhood. These queries require the LLM to produce code that invokes external APIs to answer the user's question, yet LLMs rarely produce correct code on the first try, requiring iterative code refinement upon execution results. In addition, using LLM assistants to support high query volumes can be expensive. In this work, we contribute a framework, EcoAssistant, that enables LLMs to answer code-driven queries more affordably and accurately. EcoAssistant contains three components. First, it allows the LLM assistants to converse with an automatic code executor to iteratively refine code or to produce answers based on the execution results. Second, we use a hierarchy of LLM assistants, which attempts to answer the query with weaker, cheaper LLMs before backing off to stronger, expensive ones. Third, we retrieve solutions from past successful queries as in-context demonstrations to help subsequent queries. Empirically, we show that EcoAssistant offers distinct advantages for affordability and accuracy, surpassing GPT-4 by 10 points of success rate with less than 50% of GPT-4's cost.</description><identifier>DOI: 10.48550/arxiv.2310.03046</identifier><language>eng</language><subject>Computer Science - Artificial Intelligence ; Computer Science - Software Engineering</subject><creationdate>2023-10</creationdate><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,780,885</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2310.03046$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2310.03046$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Zhang, Jieyu</creatorcontrib><creatorcontrib>Krishna, Ranjay</creatorcontrib><creatorcontrib>Awadallah, Ahmed H</creatorcontrib><creatorcontrib>Wang, Chi</creatorcontrib><title>EcoAssistant: Using LLM Assistant More Affordably and Accurately</title><description>Today, users ask Large language models (LLMs) as assistants to answer queries that require external knowledge; they ask about the weather in a specific city, about stock prices, and even about where specific locations are within their neighborhood. These queries require the LLM to produce code that invokes external APIs to answer the user's question, yet LLMs rarely produce correct code on the first try, requiring iterative code refinement upon execution results. In addition, using LLM assistants to support high query volumes can be expensive. In this work, we contribute a framework, EcoAssistant, that enables LLMs to answer code-driven queries more affordably and accurately. EcoAssistant contains three components. First, it allows the LLM assistants to converse with an automatic code executor to iteratively refine code or to produce answers based on the execution results. Second, we use a hierarchy of LLM assistants, which attempts to answer the query with weaker, cheaper LLMs before backing off to stronger, expensive ones. Third, we retrieve solutions from past successful queries as in-context demonstrations to help subsequent queries. Empirically, we show that EcoAssistant offers distinct advantages for affordability and accuracy, surpassing GPT-4 by 10 points of success rate with less than 50% of GPT-4's cost.</description><subject>Computer Science - Artificial Intelligence</subject><subject>Computer Science - Software Engineering</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNo9j81KxDAURrNxIaMP4Mq8QMf8NU1cWYbxBzq4GdflNjdXCrWVpIp9e-sorj6-szhwGLuSYmtcWYobSF_951bpFQgtjD1nd_sw1Tn3eYZxvuUvuR9fedMc-D_khylFXhNNCaEbFg4j8jqEjwRzHJYLdkYw5Hj5txt2vN8fd49F8_zwtKubAmxlC5TSokQpgqm8w1B5CkZZpVa4PuGxdBVh2RnALqJXFH2gaHUkRwGd3rDrX-0poX1P_Rukpf1JaU8p-hvggER6</recordid><startdate>20231003</startdate><enddate>20231003</enddate><creator>Zhang, Jieyu</creator><creator>Krishna, Ranjay</creator><creator>Awadallah, Ahmed H</creator><creator>Wang, Chi</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20231003</creationdate><title>EcoAssistant: Using LLM Assistant More Affordably and Accurately</title><author>Zhang, Jieyu ; Krishna, Ranjay ; Awadallah, Ahmed H ; Wang, Chi</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a676-d116d1d10c4798dc79fc42622d1ddc709d587fd5b4adbed92fe9cfe63ef8fcd83</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Computer Science - Artificial Intelligence</topic><topic>Computer Science - Software Engineering</topic><toplevel>online_resources</toplevel><creatorcontrib>Zhang, Jieyu</creatorcontrib><creatorcontrib>Krishna, Ranjay</creatorcontrib><creatorcontrib>Awadallah, Ahmed H</creatorcontrib><creatorcontrib>Wang, Chi</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Zhang, Jieyu</au><au>Krishna, Ranjay</au><au>Awadallah, Ahmed H</au><au>Wang, Chi</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>EcoAssistant: Using LLM Assistant More Affordably and Accurately</atitle><date>2023-10-03</date><risdate>2023</risdate><abstract>Today, users ask Large language models (LLMs) as assistants to answer queries that require external knowledge; they ask about the weather in a specific city, about stock prices, and even about where specific locations are within their neighborhood. These queries require the LLM to produce code that invokes external APIs to answer the user's question, yet LLMs rarely produce correct code on the first try, requiring iterative code refinement upon execution results. In addition, using LLM assistants to support high query volumes can be expensive. In this work, we contribute a framework, EcoAssistant, that enables LLMs to answer code-driven queries more affordably and accurately. EcoAssistant contains three components. First, it allows the LLM assistants to converse with an automatic code executor to iteratively refine code or to produce answers based on the execution results. Second, we use a hierarchy of LLM assistants, which attempts to answer the query with weaker, cheaper LLMs before backing off to stronger, expensive ones. Third, we retrieve solutions from past successful queries as in-context demonstrations to help subsequent queries. Empirically, we show that EcoAssistant offers distinct advantages for affordability and accuracy, surpassing GPT-4 by 10 points of success rate with less than 50% of GPT-4's cost.</abstract><doi>10.48550/arxiv.2310.03046</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2310.03046
ispartof
issn
language eng
recordid cdi_arxiv_primary_2310_03046
source arXiv.org
subjects Computer Science - Artificial Intelligence
Computer Science - Software Engineering
title EcoAssistant: Using LLM Assistant More Affordably and Accurately
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-27T01%3A48%3A43IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=EcoAssistant:%20Using%20LLM%20Assistant%20More%20Affordably%20and%20Accurately&rft.au=Zhang,%20Jieyu&rft.date=2023-10-03&rft_id=info:doi/10.48550/arxiv.2310.03046&rft_dat=%3Carxiv_GOX%3E2310_03046%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true