Learning to Communicate with Strangers via Channel Randomisation Methods

Learning to Communicate with Strangers via Channel Randomisation Methods

4th Workshop on Emergent Communication at NeurIPS 2020 We introduce two methods for improving the performance of agents meeting for the first time to accomplish a communicative task. The methods are: (1) `message mutation' during the generation of the communication protocol; and (2) random perm...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Cope, Dylan, Schoots, Nandi
Format: Artikel
Sprache:eng
Schlagworte:
Computer Science - Artificial Intelligence
Computer Science - Computation and Language
Computer Science - Learning
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 Cope, Dylan
Schoots, Nandi
description 4th Workshop on Emergent Communication at NeurIPS 2020 We introduce two methods for improving the performance of agents meeting for the first time to accomplish a communicative task. The methods are: (1) `message mutation' during the generation of the communication protocol; and (2) random permutations of the communication channel. These proposals are tested using a simple two-player game involving a `teacher' who generates a communication protocol and sends a message, and a `student' who interprets the message. After training multiple agents via self-play we analyse the performance of these agents when they are matched with a stranger, i.e. their zero-shot communication performance. We find that both message mutation and channel permutation positively influence performance, and we discuss their effects.
doi_str_mv 10.48550/arxiv.2104.09557
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2104_09557</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2104_09557</sourcerecordid><originalsourceid>FETCH-LOGICAL-a677-4cdc6de124a1389924c6947fae5dd806879acfe0fe3a234f4d22ca59b7c0ba13</originalsourceid><addsrcrecordid>eNotz7FOwzAUhWEvDKjtAzDhF0hwHDuOxyqiLVIQEmWPbu3rxlJjI8ct8PZAYTrTf6SPkLuKlaKVkj1A-vSXkldMlExLqW7JrkdIwYcjzZF2cZrOwRvISD98Huk-JwhHTDO9eKDdCCHgib5CsHHyM2QfA33GPEY7L8mNg9OMq_9dkP3m8a3bFf3L9qlb9wU0ShXCWNNYrLiAqm615sI0WigHKK1tWdMqDcYhc1gDr4UTlnMDUh-UYYefZEHu_16vkuE9-QnS1_ArGq6i-huu60dT</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Learning to Communicate with Strangers via Channel Randomisation Methods</title><source>arXiv.org</source><creator>Cope, Dylan ; Schoots, Nandi</creator><creatorcontrib>Cope, Dylan ; Schoots, Nandi</creatorcontrib><description>4th Workshop on Emergent Communication at NeurIPS 2020 We introduce two methods for improving the performance of agents meeting for the first time to accomplish a communicative task. The methods are: (1) `message mutation' during the generation of the communication protocol; and (2) random permutations of the communication channel. These proposals are tested using a simple two-player game involving a `teacher' who generates a communication protocol and sends a message, and a `student' who interprets the message. After training multiple agents via self-play we analyse the performance of these agents when they are matched with a stranger, i.e. their zero-shot communication performance. We find that both message mutation and channel permutation positively influence performance, and we discuss their effects.</description><identifier>DOI: 10.48550/arxiv.2104.09557</identifier><language>eng</language><subject>Computer Science - Artificial Intelligence ; Computer Science - Computation and Language ; Computer Science - Learning</subject><creationdate>2021-04</creationdate><rights>http://creativecommons.org/licenses/by/4.0</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/2104.09557$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2104.09557$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Cope, Dylan</creatorcontrib><creatorcontrib>Schoots, Nandi</creatorcontrib><title>Learning to Communicate with Strangers via Channel Randomisation Methods</title><description>4th Workshop on Emergent Communication at NeurIPS 2020 We introduce two methods for improving the performance of agents meeting for the first time to accomplish a communicative task. The methods are: (1) `message mutation' during the generation of the communication protocol; and (2) random permutations of the communication channel. These proposals are tested using a simple two-player game involving a `teacher' who generates a communication protocol and sends a message, and a `student' who interprets the message. After training multiple agents via self-play we analyse the performance of these agents when they are matched with a stranger, i.e. their zero-shot communication performance. We find that both message mutation and channel permutation positively influence performance, and we discuss their effects.</description><subject>Computer Science - Artificial Intelligence</subject><subject>Computer Science - Computation and Language</subject><subject>Computer Science - Learning</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotz7FOwzAUhWEvDKjtAzDhF0hwHDuOxyqiLVIQEmWPbu3rxlJjI8ct8PZAYTrTf6SPkLuKlaKVkj1A-vSXkldMlExLqW7JrkdIwYcjzZF2cZrOwRvISD98Huk-JwhHTDO9eKDdCCHgib5CsHHyM2QfA33GPEY7L8mNg9OMq_9dkP3m8a3bFf3L9qlb9wU0ShXCWNNYrLiAqm615sI0WigHKK1tWdMqDcYhc1gDr4UTlnMDUh-UYYefZEHu_16vkuE9-QnS1_ArGq6i-huu60dT</recordid><startdate>20210419</startdate><enddate>20210419</enddate><creator>Cope, Dylan</creator><creator>Schoots, Nandi</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20210419</creationdate><title>Learning to Communicate with Strangers via Channel Randomisation Methods</title><author>Cope, Dylan ; Schoots, Nandi</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a677-4cdc6de124a1389924c6947fae5dd806879acfe0fe3a234f4d22ca59b7c0ba13</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Computer Science - Artificial Intelligence</topic><topic>Computer Science - Computation and Language</topic><topic>Computer Science - Learning</topic><toplevel>online_resources</toplevel><creatorcontrib>Cope, Dylan</creatorcontrib><creatorcontrib>Schoots, Nandi</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Cope, Dylan</au><au>Schoots, Nandi</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Learning to Communicate with Strangers via Channel Randomisation Methods</atitle><date>2021-04-19</date><risdate>2021</risdate><abstract>4th Workshop on Emergent Communication at NeurIPS 2020 We introduce two methods for improving the performance of agents meeting for the first time to accomplish a communicative task. The methods are: (1) `message mutation' during the generation of the communication protocol; and (2) random permutations of the communication channel. These proposals are tested using a simple two-player game involving a `teacher' who generates a communication protocol and sends a message, and a `student' who interprets the message. After training multiple agents via self-play we analyse the performance of these agents when they are matched with a stranger, i.e. their zero-shot communication performance. We find that both message mutation and channel permutation positively influence performance, and we discuss their effects.</abstract><doi>10.48550/arxiv.2104.09557</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2104.09557
ispartof
issn
language eng
recordid cdi_arxiv_primary_2104_09557
source arXiv.org
subjects Computer Science - Artificial Intelligence
Computer Science - Computation and Language
Computer Science - Learning
title Learning to Communicate with Strangers via Channel Randomisation Methods
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-08T04%3A47%3A30IST&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=Learning%20to%20Communicate%20with%20Strangers%20via%20Channel%20Randomisation%20Methods&rft.au=Cope,%20Dylan&rft.date=2021-04-19&rft_id=info:doi/10.48550/arxiv.2104.09557&rft_dat=%3Carxiv_GOX%3E2104_09557%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