Matches in SemOpenAlex for { <https://semopenalex.org/work/W4378711560> ?p ?o ?g. }
Showing items 1 to 85 of
85
with 100 items per page.
- W4378711560 abstract "A self-explaining rationalization model is generally constructed by a cooperative game where a generator selects the most human-intelligible pieces from the input text as rationales, followed by a predictor that makes predictions based on the selected rationales. However, such a cooperative game may incur the degeneration problem where the predictor overfits to the uninformative pieces generated by a not yet well-trained generator and in turn, leads the generator to converge to a sub-optimal model that tends to select senseless pieces. In this paper, we theoretically bridge degeneration with the predictor's Lipschitz continuity. Then, we empirically propose a simple but effective method named DR, which can naturally and flexibly restrain the Lipschitz constant of the predictor, to address the problem of degeneration. The main idea of DR is to decouple the generator and predictor to allocate them with asymmetric learning rates. A series of experiments conducted on two widely used benchmarks have verified the effectiveness of the proposed method. Codes: https://github.com/jugechengzi/Rationalization-DR." @default.
- W4378711560 created "2023-05-30" @default.
- W4378711560 creator A5008831642 @default.
- W4378711560 creator A5027119797 @default.
- W4378711560 creator A5037997848 @default.
- W4378711560 creator A5042241049 @default.
- W4378711560 creator A5071037763 @default.
- W4378711560 creator A5075195203 @default.
- W4378711560 creator A5076460648 @default.
- W4378711560 creator A5092046693 @default.
- W4378711560 date "2023-08-04" @default.
- W4378711560 modified "2023-09-27" @default.
- W4378711560 title "Decoupled Rationalization with Asymmetric Learning Rates: A Flexible Lipschitz Restraint" @default.
- W4378711560 cites W2001259128 @default.
- W4378711560 cites W2157331557 @default.
- W4378711560 cites W2160409620 @default.
- W4378711560 cites W2889436406 @default.
- W4378711560 cites W2918073309 @default.
- W4378711560 cites W2949227999 @default.
- W4378711560 cites W2963233086 @default.
- W4378711560 cites W2970155250 @default.
- W4378711560 cites W2997072274 @default.
- W4378711560 cites W3035064231 @default.
- W4378711560 cites W3105868192 @default.
- W4378711560 cites W3165967177 @default.
- W4378711560 cites W3206945533 @default.
- W4378711560 cites W4224950688 @default.
- W4378711560 cites W4242847488 @default.
- W4378711560 doi "https://doi.org/10.1145/3580305.3599299" @default.
- W4378711560 hasPublicationYear "2023" @default.
- W4378711560 type Work @default.
- W4378711560 citedByCount "0" @default.
- W4378711560 crossrefType "proceedings-article" @default.
- W4378711560 hasAuthorship W4378711560A5008831642 @default.
- W4378711560 hasAuthorship W4378711560A5027119797 @default.
- W4378711560 hasAuthorship W4378711560A5037997848 @default.
- W4378711560 hasAuthorship W4378711560A5042241049 @default.
- W4378711560 hasAuthorship W4378711560A5071037763 @default.
- W4378711560 hasAuthorship W4378711560A5075195203 @default.
- W4378711560 hasAuthorship W4378711560A5076460648 @default.
- W4378711560 hasAuthorship W4378711560A5092046693 @default.
- W4378711560 hasBestOaLocation W43787115601 @default.
- W4378711560 hasConcept C121332964 @default.
- W4378711560 hasConcept C134306372 @default.
- W4378711560 hasConcept C154945302 @default.
- W4378711560 hasConcept C162324750 @default.
- W4378711560 hasConcept C163258240 @default.
- W4378711560 hasConcept C175444787 @default.
- W4378711560 hasConcept C22324862 @default.
- W4378711560 hasConcept C2780992000 @default.
- W4378711560 hasConcept C33923547 @default.
- W4378711560 hasConcept C41008148 @default.
- W4378711560 hasConcept C52438962 @default.
- W4378711560 hasConcept C62520636 @default.
- W4378711560 hasConceptScore W4378711560C121332964 @default.
- W4378711560 hasConceptScore W4378711560C134306372 @default.
- W4378711560 hasConceptScore W4378711560C154945302 @default.
- W4378711560 hasConceptScore W4378711560C162324750 @default.
- W4378711560 hasConceptScore W4378711560C163258240 @default.
- W4378711560 hasConceptScore W4378711560C175444787 @default.
- W4378711560 hasConceptScore W4378711560C22324862 @default.
- W4378711560 hasConceptScore W4378711560C2780992000 @default.
- W4378711560 hasConceptScore W4378711560C33923547 @default.
- W4378711560 hasConceptScore W4378711560C41008148 @default.
- W4378711560 hasConceptScore W4378711560C52438962 @default.
- W4378711560 hasConceptScore W4378711560C62520636 @default.
- W4378711560 hasFunder F4320321001 @default.
- W4378711560 hasLocation W43787115601 @default.
- W4378711560 hasLocation W43787115602 @default.
- W4378711560 hasLocation W43787115603 @default.
- W4378711560 hasOpenAccess W4378711560 @default.
- W4378711560 hasPrimaryLocation W43787115601 @default.
- W4378711560 hasRelatedWork W2052539338 @default.
- W4378711560 hasRelatedWork W2062740134 @default.
- W4378711560 hasRelatedWork W2356629573 @default.
- W4378711560 hasRelatedWork W2390459957 @default.
- W4378711560 hasRelatedWork W2746742710 @default.
- W4378711560 hasRelatedWork W2949940795 @default.
- W4378711560 hasRelatedWork W2953067553 @default.
- W4378711560 hasRelatedWork W4289422074 @default.
- W4378711560 hasRelatedWork W4289422133 @default.
- W4378711560 hasRelatedWork W4378711560 @default.
- W4378711560 isParatext "false" @default.
- W4378711560 isRetracted "false" @default.
- W4378711560 workType "article" @default.