Matches in SemOpenAlex for { <https://semopenalex.org/work/W2805862167> ?p ?o ?g. }
Showing items 1 to 88 of
88
with 100 items per page.
- W2805862167 abstract "How can we design reinforcement learning agents that avoid causing unnecessary disruptions to their environment? We argue that current approaches to penalizing side effects can introduce bad incentives in tasks that require irreversible actions, and in environments that contain sources of change other than the agent. For example, some approaches give the agent an incentive to prevent any irreversible changes in the environment, including the actions of other agents. We introduce a general definition of side effects, based on relative reachability of states compared to a default state, that avoids these undesirable incentives. Using a set of gridworld experiments illustrating relevant scenarios, we empirically compare relative reachability to penalties based on existing definitions and show that it is the only penalty among those tested that produces the desired behavior in all the scenarios." @default.
- W2805862167 created "2018-06-13" @default.
- W2805862167 creator A5008987732 @default.
- W2805862167 creator A5025651820 @default.
- W2805862167 creator A5052300917 @default.
- W2805862167 creator A5081226725 @default.
- W2805862167 date "2018-06-04" @default.
- W2805862167 modified "2023-09-27" @default.
- W2805862167 title "Measuring and avoiding side effects using relative reachability" @default.
- W2805862167 cites W1585575029 @default.
- W2805862167 cites W1786044565 @default.
- W2805862167 cites W1963790880 @default.
- W2805862167 cites W1999874108 @default.
- W2805862167 cites W2061562262 @default.
- W2805862167 cites W2098774185 @default.
- W2805862167 cites W2116364955 @default.
- W2805862167 cites W2288565641 @default.
- W2805862167 cites W2626804490 @default.
- W2805862167 cites W2750866482 @default.
- W2805862167 cites W2952720101 @default.
- W2805862167 cites W2962730405 @default.
- W2805862167 cites W2963289505 @default.
- W2805862167 cites W2963489214 @default.
- W2805862167 hasPublicationYear "2018" @default.
- W2805862167 type Work @default.
- W2805862167 sameAs 2805862167 @default.
- W2805862167 citedByCount "8" @default.
- W2805862167 countsByYear W28058621672018 @default.
- W2805862167 countsByYear W28058621672019 @default.
- W2805862167 countsByYear W28058621672020 @default.
- W2805862167 crossrefType "posted-content" @default.
- W2805862167 hasAuthorship W2805862167A5008987732 @default.
- W2805862167 hasAuthorship W2805862167A5025651820 @default.
- W2805862167 hasAuthorship W2805862167A5052300917 @default.
- W2805862167 hasAuthorship W2805862167A5081226725 @default.
- W2805862167 hasConcept C112930515 @default.
- W2805862167 hasConcept C11413529 @default.
- W2805862167 hasConcept C136643341 @default.
- W2805862167 hasConcept C144133560 @default.
- W2805862167 hasConcept C154945302 @default.
- W2805862167 hasConcept C162324750 @default.
- W2805862167 hasConcept C175444787 @default.
- W2805862167 hasConcept C177264268 @default.
- W2805862167 hasConcept C199360897 @default.
- W2805862167 hasConcept C29122968 @default.
- W2805862167 hasConcept C41008148 @default.
- W2805862167 hasConcept C48103436 @default.
- W2805862167 hasConcept C97541855 @default.
- W2805862167 hasConceptScore W2805862167C112930515 @default.
- W2805862167 hasConceptScore W2805862167C11413529 @default.
- W2805862167 hasConceptScore W2805862167C136643341 @default.
- W2805862167 hasConceptScore W2805862167C144133560 @default.
- W2805862167 hasConceptScore W2805862167C154945302 @default.
- W2805862167 hasConceptScore W2805862167C162324750 @default.
- W2805862167 hasConceptScore W2805862167C175444787 @default.
- W2805862167 hasConceptScore W2805862167C177264268 @default.
- W2805862167 hasConceptScore W2805862167C199360897 @default.
- W2805862167 hasConceptScore W2805862167C29122968 @default.
- W2805862167 hasConceptScore W2805862167C41008148 @default.
- W2805862167 hasConceptScore W2805862167C48103436 @default.
- W2805862167 hasConceptScore W2805862167C97541855 @default.
- W2805862167 hasLocation W28058621671 @default.
- W2805862167 hasOpenAccess W2805862167 @default.
- W2805862167 hasPrimaryLocation W28058621671 @default.
- W2805862167 hasRelatedWork W1591912923 @default.
- W2805862167 hasRelatedWork W2084270684 @default.
- W2805862167 hasRelatedWork W2145339207 @default.
- W2805862167 hasRelatedWork W2186794832 @default.
- W2805862167 hasRelatedWork W2253998085 @default.
- W2805862167 hasRelatedWork W2398981685 @default.
- W2805862167 hasRelatedWork W2401180666 @default.
- W2805862167 hasRelatedWork W2462906003 @default.
- W2805862167 hasRelatedWork W2617191410 @default.
- W2805862167 hasRelatedWork W2761566394 @default.
- W2805862167 hasRelatedWork W2768908787 @default.
- W2805862167 hasRelatedWork W2788578887 @default.
- W2805862167 hasRelatedWork W2883614491 @default.
- W2805862167 hasRelatedWork W2899825019 @default.
- W2805862167 hasRelatedWork W2963193690 @default.
- W2805862167 hasRelatedWork W3016025770 @default.
- W2805862167 hasRelatedWork W3097059970 @default.
- W2805862167 hasRelatedWork W3117850269 @default.
- W2805862167 hasRelatedWork W3173700155 @default.
- W2805862167 hasRelatedWork W399024592 @default.
- W2805862167 isParatext "false" @default.
- W2805862167 isRetracted "false" @default.
- W2805862167 magId "2805862167" @default.
- W2805862167 workType "article" @default.