Matches in SemOpenAlex for { <https://semopenalex.org/work/W3084850341> ?p ?o ?g. }
- W3084850341 abstract "Existing open-domain dialog models are generally trained to minimize the perplexity of target human responses. However, some human replies are more engaging than others, spawning more followup interactions. Current conversational models are increasingly capable of producing turns that are context-relevant, but in order to produce compelling agents, these models need to be able to predict and optimize for turns that are genuinely engaging. We leverage social media feedback data (number of replies and upvotes) to build a large-scale training dataset for feedback prediction. To alleviate possible distortion between the feedback and engagingness, we convert the ranking problem to a comparison of response pairs which involve few confounding factors. We trained DialogRPT, a set of GPT-2 based models on 133M pairs of human feedback data and the resulting ranker outperformed several baselines. Particularly, our ranker outperforms the conventional dialog perplexity baseline with a large margin on predicting Reddit feedback. We finally combine the feedback prediction models and a human-like scoring model to rank the machine-generated dialog responses. Crowd-sourced human evaluation shows that our ranking method correlates better with real human preferences than baseline models." @default.
- W3084850341 created "2020-09-21" @default.
- W3084850341 creator A5015656536 @default.
- W3084850341 creator A5037066465 @default.
- W3084850341 creator A5057693146 @default.
- W3084850341 creator A5075372122 @default.
- W3084850341 creator A5081872106 @default.
- W3084850341 date "2020-09-15" @default.
- W3084850341 modified "2023-09-25" @default.
- W3084850341 title "Dialogue Response Ranking Training with Large-Scale Human Feedback Data" @default.
- W3084850341 cites W105848778 @default.
- W3084850341 cites W2019502441 @default.
- W3084850341 cites W2101105183 @default.
- W3084850341 cites W2130041324 @default.
- W3084850341 cites W2138621090 @default.
- W3084850341 cites W2147453867 @default.
- W3084850341 cites W2155482025 @default.
- W3084850341 cites W2264742718 @default.
- W3084850341 cites W2605246398 @default.
- W3084850341 cites W2761590056 @default.
- W3084850341 cites W2784400615 @default.
- W3084850341 cites W2906579211 @default.
- W3084850341 cites W2916898195 @default.
- W3084850341 cites W2936695845 @default.
- W3084850341 cites W2949555952 @default.
- W3084850341 cites W2951883832 @default.
- W3084850341 cites W2953206424 @default.
- W3084850341 cites W2963206148 @default.
- W3084850341 cites W2963667505 @default.
- W3084850341 cites W2969574947 @default.
- W3084850341 cites W2970785611 @default.
- W3084850341 cites W2988299267 @default.
- W3084850341 cites W3000779003 @default.
- W3084850341 cites W3005680577 @default.
- W3084850341 cites W3016254928 @default.
- W3084850341 cites W3023786569 @default.
- W3084850341 cites W3024436954 @default.
- W3084850341 cites W3037026762 @default.
- W3084850341 doi "https://doi.org/10.48550/arxiv.2009.06978" @default.
- W3084850341 hasPublicationYear "2020" @default.
- W3084850341 type Work @default.
- W3084850341 sameAs 3084850341 @default.
- W3084850341 citedByCount "5" @default.
- W3084850341 countsByYear W30848503412021 @default.
- W3084850341 countsByYear W30848503412022 @default.
- W3084850341 crossrefType "posted-content" @default.
- W3084850341 hasAuthorship W3084850341A5015656536 @default.
- W3084850341 hasAuthorship W3084850341A5037066465 @default.
- W3084850341 hasAuthorship W3084850341A5057693146 @default.
- W3084850341 hasAuthorship W3084850341A5075372122 @default.
- W3084850341 hasAuthorship W3084850341A5081872106 @default.
- W3084850341 hasBestOaLocation W30848503411 @default.
- W3084850341 hasConcept C100279451 @default.
- W3084850341 hasConcept C111368507 @default.
- W3084850341 hasConcept C119857082 @default.
- W3084850341 hasConcept C12725497 @default.
- W3084850341 hasConcept C127313418 @default.
- W3084850341 hasConcept C136764020 @default.
- W3084850341 hasConcept C137293760 @default.
- W3084850341 hasConcept C151730666 @default.
- W3084850341 hasConcept C153083717 @default.
- W3084850341 hasConcept C154945302 @default.
- W3084850341 hasConcept C173853756 @default.
- W3084850341 hasConcept C189430467 @default.
- W3084850341 hasConcept C2779343474 @default.
- W3084850341 hasConcept C41008148 @default.
- W3084850341 hasConcept C51632099 @default.
- W3084850341 hasConcept C86803240 @default.
- W3084850341 hasConceptScore W3084850341C100279451 @default.
- W3084850341 hasConceptScore W3084850341C111368507 @default.
- W3084850341 hasConceptScore W3084850341C119857082 @default.
- W3084850341 hasConceptScore W3084850341C12725497 @default.
- W3084850341 hasConceptScore W3084850341C127313418 @default.
- W3084850341 hasConceptScore W3084850341C136764020 @default.
- W3084850341 hasConceptScore W3084850341C137293760 @default.
- W3084850341 hasConceptScore W3084850341C151730666 @default.
- W3084850341 hasConceptScore W3084850341C153083717 @default.
- W3084850341 hasConceptScore W3084850341C154945302 @default.
- W3084850341 hasConceptScore W3084850341C173853756 @default.
- W3084850341 hasConceptScore W3084850341C189430467 @default.
- W3084850341 hasConceptScore W3084850341C2779343474 @default.
- W3084850341 hasConceptScore W3084850341C41008148 @default.
- W3084850341 hasConceptScore W3084850341C51632099 @default.
- W3084850341 hasConceptScore W3084850341C86803240 @default.
- W3084850341 hasLocation W30848503411 @default.
- W3084850341 hasOpenAccess W3084850341 @default.
- W3084850341 hasPrimaryLocation W30848503411 @default.
- W3084850341 hasRelatedWork W10980763 @default.
- W3084850341 hasRelatedWork W11223206 @default.
- W3084850341 hasRelatedWork W13303079 @default.
- W3084850341 hasRelatedWork W133296 @default.
- W3084850341 hasRelatedWork W14204599 @default.
- W3084850341 hasRelatedWork W2882640 @default.
- W3084850341 hasRelatedWork W2990258 @default.
- W3084850341 hasRelatedWork W6310906 @default.
- W3084850341 hasRelatedWork W6915741 @default.
- W3084850341 hasRelatedWork W8444177 @default.
- W3084850341 isParatext "false" @default.
- W3084850341 isRetracted "false" @default.
- W3084850341 magId "3084850341" @default.