FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3197979077> ?p ?o ?g. }

• W3197979077 abstract "Recent prompt-based approaches allow pretrained language models to achieve strong performances on few-shot finetuning by reformulating downstream tasks as a language modeling problem. In this work, we demonstrate that, despite its advantages on low data regimes, finetuned prompt-based models for sentence pair classification tasks still suffer from a common pitfall of adopting inference heuristics based on lexical overlap, e.g., models incorrectly assuming a sentence pair is of the same meaning because they consist of the same set of words. Interestingly, we find that this particular inference heuristic is significantly less present in the zero-shot evaluation of the prompt-based model, indicating how finetuning can be destructive to useful knowledge learned during the pretraining. We then show that adding a regularization that preserves pretraining weights is effective in mitigating this destructive tendency of few-shot finetuning. Our evaluation on three datasets demonstrates promising improvements on the three corresponding challenge datasets used to diagnose the inference heuristics." @default.
• W3197979077 created "2021-09-13" @default.
• W3197979077 creator A5027450194 @default.
• W3197979077 creator A5040728796 @default.
• W3197979077 creator A5049805631 @default.
• W3197979077 creator A5054918343 @default.
• W3197979077 date "2021-01-01" @default.
• W3197979077 modified "2023-09-30" @default.
• W3197979077 title "Avoiding Inference Heuristics in Few-shot Prompt-based Finetuning" @default.
• W3197979077 cites W1821462560 @default.
• W3197979077 cites W1840435438 @default.
• W3197979077 cites W2060277733 @default.
• W3197979077 cites W2560647685 @default.
• W3197979077 cites W2739505524 @default.
• W3197979077 cites W2798665661 @default.
• W3197979077 cites W2932893307 @default.
• W3197979077 cites W2946359678 @default.
• W3197979077 cites W2946417913 @default.
• W3197979077 cites W2946609015 @default.
• W3197979077 cites W2951286828 @default.
• W3197979077 cites W2952984539 @default.
• W3197979077 cites W2962727366 @default.
• W3197979077 cites W2962843521 @default.
• W3197979077 cites W2963120843 @default.
• W3197979077 cites W2963122961 @default.
• W3197979077 cites W2963310665 @default.
• W3197979077 cites W2963372062 @default.
• W3197979077 cites W2963383094 @default.
• W3197979077 cites W2963394326 @default.
• W3197979077 cites W2963542100 @default.
• W3197979077 cites W2963674932 @default.
• W3197979077 cites W2963846996 @default.
• W3197979077 cites W2963969878 @default.
• W3197979077 cites W2964150944 @default.
• W3197979077 cites W2965373594 @default.
• W3197979077 cites W2970379526 @default.
• W3197979077 cites W2978017171 @default.
• W3197979077 cites W2983984338 @default.
• W3197979077 cites W2984256198 @default.
• W3197979077 cites W2994934025 @default.
• W3197979077 cites W2996908057 @default.
• W3197979077 cites W3005700362 @default.
• W3197979077 cites W3034238904 @default.
• W3197979077 cites W3034292689 @default.
• W3197979077 cites W3034831508 @default.
• W3197979077 cites W3034850762 @default.
• W3197979077 cites W3035139434 @default.
• W3197979077 cites W3035352537 @default.
• W3197979077 cites W3086499488 @default.
• W3197979077 cites W3091818438 @default.
• W3197979077 cites W3093655762 @default.
• W3197979077 cites W3098613713 @default.
• W3197979077 cites W3100895823 @default.
• W3197979077 cites W3102259594 @default.
• W3197979077 cites W3104215796 @default.
• W3197979077 cites W3104738015 @default.
• W3197979077 cites W3115772171 @default.
• W3197979077 cites W3120490999 @default.
• W3197979077 cites W3126493605 @default.
• W3197979077 cites W3137585426 @default.
• W3197979077 cites W3152911627 @default.
• W3197979077 cites W3153427360 @default.
• W3197979077 cites W3167602185 @default.
• W3197979077 cites W3170180819 @default.
• W3197979077 cites W3172642864 @default.
• W3197979077 cites W3173777717 @default.
• W3197979077 cites W3214897310 @default.
• W3197979077 doi "https://doi.org/10.18653/v1/2021.emnlp-main.713" @default.
• W3197979077 hasPublicationYear "2021" @default.
• W3197979077 type Work @default.
• W3197979077 sameAs 3197979077 @default.
• W3197979077 citedByCount "1" @default.
• W3197979077 countsByYear W31979790772023 @default.
• W3197979077 crossrefType "proceedings-article" @default.
• W3197979077 hasAuthorship W3197979077A5027450194 @default.
• W3197979077 hasAuthorship W3197979077A5040728796 @default.
• W3197979077 hasAuthorship W3197979077A5049805631 @default.
• W3197979077 hasAuthorship W3197979077A5054918343 @default.
• W3197979077 hasBestOaLocation W31979790771 @default.
• W3197979077 hasConcept C111919701 @default.
• W3197979077 hasConcept C119857082 @default.
• W3197979077 hasConcept C127705205 @default.
• W3197979077 hasConcept C137293760 @default.
• W3197979077 hasConcept C154945302 @default.
• W3197979077 hasConcept C173801870 @default.
• W3197979077 hasConcept C177264268 @default.
• W3197979077 hasConcept C199360897 @default.
• W3197979077 hasConcept C204321447 @default.
• W3197979077 hasConcept C2776214188 @default.
• W3197979077 hasConcept C2777530160 @default.
• W3197979077 hasConcept C41008148 @default.
• W3197979077 hasConceptScore W3197979077C111919701 @default.
• W3197979077 hasConceptScore W3197979077C119857082 @default.
• W3197979077 hasConceptScore W3197979077C127705205 @default.
• W3197979077 hasConceptScore W3197979077C137293760 @default.
• W3197979077 hasConceptScore W3197979077C154945302 @default.
• W3197979077 hasConceptScore W3197979077C173801870 @default.
• W3197979077 hasConceptScore W3197979077C177264268 @default.
• W3197979077 hasConceptScore W3197979077C199360897 @default.
• W3197979077 hasConceptScore W3197979077C204321447 @default.

• next