Measuring Goal-Directedness

We define maximum entropy goal-directedness (MEG), a formal measure of goal-directedness in causal models and Markov decision processes, and give algorithms for computing it. Measuring goal-directedness is important, as it is a critical element of many concerns about harm from AI. It is also of phil...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2024-12
Main Authors: MacDermott, Matt, Fox, James, Belardinelli, Francesco, Everitt, Tom
Format: Article
Language:English
Subjects:
Algorithms
Machine learning
Markov processes
Maximum entropy
Random variables
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by
cites
container_end_page
container_issue
container_start_page
container_title arXiv.org
container_volume
creator MacDermott, Matt
Fox, James
Belardinelli, Francesco
Everitt, Tom
description We define maximum entropy goal-directedness (MEG), a formal measure of goal-directedness in causal models and Markov decision processes, and give algorithms for computing it. Measuring goal-directedness is important, as it is a critical element of many concerns about harm from AI. It is also of philosophical interest, as goal-directedness is a key aspect of agency. MEG is based on an adaptation of the maximum causal entropy framework used in inverse reinforcement learning. It can measure goal-directedness with respect to a known utility function, a hypothesis class of utility functions, or a set of random variables. We prove that MEG satisfies several desiderata and demonstrate our algorithms with small-scale experiments.
format article
fullrecord <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_3142376045</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>3142376045</sourcerecordid><originalsourceid>FETCH-proquest_journals_31423760453</originalsourceid><addsrcrecordid>eNpjYuA0MjY21LUwMTLiYOAtLs4yMDAwMjM3MjU15mSQ9k1NLC4tysxLV3DPT8zRdcksSk0uSU3JSy0u5mFgTUvMKU7lhdLcDMpuriHOHroFRfmFpanFJfFZ-aVFeUCpeGNDEyNjczMDE1Nj4lQBAPsxKz4</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>3142376045</pqid></control><display><type>article</type><title>Measuring Goal-Directedness</title><source>Publicly Available Content Database (Proquest) (PQ_SDU_P3)</source><creator>MacDermott, Matt ; Fox, James ; Belardinelli, Francesco ; Everitt, Tom</creator><creatorcontrib>MacDermott, Matt ; Fox, James ; Belardinelli, Francesco ; Everitt, Tom</creatorcontrib><description>We define maximum entropy goal-directedness (MEG), a formal measure of goal-directedness in causal models and Markov decision processes, and give algorithms for computing it. Measuring goal-directedness is important, as it is a critical element of many concerns about harm from AI. It is also of philosophical interest, as goal-directedness is a key aspect of agency. MEG is based on an adaptation of the maximum causal entropy framework used in inverse reinforcement learning. It can measure goal-directedness with respect to a known utility function, a hypothesis class of utility functions, or a set of random variables. We prove that MEG satisfies several desiderata and demonstrate our algorithms with small-scale experiments.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Algorithms ; Machine learning ; Markov processes ; Maximum entropy ; Random variables</subject><ispartof>arXiv.org, 2024-12</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>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://www.proquest.com/docview/3142376045?pq-origsite=primo$$EHTML$$P50$$Gproquest$$Hfree_for_read</linktohtml><link.rule.ids>780,784,25753,37012,44590</link.rule.ids></links><search><creatorcontrib>MacDermott, Matt</creatorcontrib><creatorcontrib>Fox, James</creatorcontrib><creatorcontrib>Belardinelli, Francesco</creatorcontrib><creatorcontrib>Everitt, Tom</creatorcontrib><title>Measuring Goal-Directedness</title><title>arXiv.org</title><description>We define maximum entropy goal-directedness (MEG), a formal measure of goal-directedness in causal models and Markov decision processes, and give algorithms for computing it. Measuring goal-directedness is important, as it is a critical element of many concerns about harm from AI. It is also of philosophical interest, as goal-directedness is a key aspect of agency. MEG is based on an adaptation of the maximum causal entropy framework used in inverse reinforcement learning. It can measure goal-directedness with respect to a known utility function, a hypothesis class of utility functions, or a set of random variables. We prove that MEG satisfies several desiderata and demonstrate our algorithms with small-scale experiments.</description><subject>Algorithms</subject><subject>Machine learning</subject><subject>Markov processes</subject><subject>Maximum entropy</subject><subject>Random variables</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>PIMPY</sourceid><recordid>eNpjYuA0MjY21LUwMTLiYOAtLs4yMDAwMjM3MjU15mSQ9k1NLC4tysxLV3DPT8zRdcksSk0uSU3JSy0u5mFgTUvMKU7lhdLcDMpuriHOHroFRfmFpanFJfFZ-aVFeUCpeGNDEyNjczMDE1Nj4lQBAPsxKz4</recordid><startdate>20241206</startdate><enddate>20241206</enddate><creator>MacDermott, Matt</creator><creator>Fox, James</creator><creator>Belardinelli, Francesco</creator><creator>Everitt, Tom</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>20241206</creationdate><title>Measuring Goal-Directedness</title><author>MacDermott, Matt ; Fox, James ; Belardinelli, Francesco ; Everitt, Tom</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_31423760453</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Algorithms</topic><topic>Machine learning</topic><topic>Markov processes</topic><topic>Maximum entropy</topic><topic>Random variables</topic><toplevel>online_resources</toplevel><creatorcontrib>MacDermott, Matt</creatorcontrib><creatorcontrib>Fox, James</creatorcontrib><creatorcontrib>Belardinelli, Francesco</creatorcontrib><creatorcontrib>Everitt, Tom</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science &amp; Engineering Collection</collection><collection>ProQuest Central (Alumni)</collection><collection>ProQuest Central</collection><collection>ProQuest Central Essentials</collection><collection>AUTh Library subscriptions: ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central</collection><collection>SciTech Premium Collection (Proquest) (PQ_SDU_P3)</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database (Proquest) (PQ_SDU_P3)</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>MacDermott, Matt</au><au>Fox, James</au><au>Belardinelli, Francesco</au><au>Everitt, Tom</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>Measuring Goal-Directedness</atitle><jtitle>arXiv.org</jtitle><date>2024-12-06</date><risdate>2024</risdate><eissn>2331-8422</eissn><abstract>We define maximum entropy goal-directedness (MEG), a formal measure of goal-directedness in causal models and Markov decision processes, and give algorithms for computing it. Measuring goal-directedness is important, as it is a critical element of many concerns about harm from AI. It is also of philosophical interest, as goal-directedness is a key aspect of agency. MEG is based on an adaptation of the maximum causal entropy framework used in inverse reinforcement learning. It can measure goal-directedness with respect to a known utility function, a hypothesis class of utility functions, or a set of random variables. We prove that MEG satisfies several desiderata and demonstrate our algorithms with small-scale experiments.</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, 2024-12
issn 2331-8422
language eng
recordid cdi_proquest_journals_3142376045
source Publicly Available Content Database (Proquest) (PQ_SDU_P3)
subjects Algorithms
Machine learning
Markov processes
Maximum entropy
Random variables
title Measuring Goal-Directedness
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-02T15%3A09%3A10IST&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=Measuring%20Goal-Directedness&rft.jtitle=arXiv.org&rft.au=MacDermott,%20Matt&rft.date=2024-12-06&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E3142376045%3C/proquest%3E%3Cgrp_id%3Ecdi_FETCH-proquest_journals_31423760453%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_pqid=3142376045&rft_id=info:pmid/&rfr_iscdi=true