AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

DARE: Disentanglement-Augmented Rationale Extraction

Linan Yue, Qi Liu, Yichao Du, Yanqing An, Li Wang, Enhong Chen

NeurIPS 2022 · Label-efficient and robust learning › Recognition and long-tail · approach: Dense features and foundation priors · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Add disentanglement of non-rationale tokens with MI minimization (HardKuma -> DARE) (ablation)HardKuma · BeerAdvocate Appearancetoken F1: 78.3 -> 82.9Improved
Add disentanglement of non-rationale tokens with MI minimization (HardKuma -> DARE) (ablation)HardKuma · BeerAdvocate Smelltoken F1: 47.5 -> 61.2Improved
Add disentanglement of non-rationale tokens with MI minimization (HardKuma -> DARE) (ablation)HardKuma · BeerAdvocate Palatetoken F1: 63.1 -> 69.9Improved
Replace CLUB_NCE with L1Out as MI minimizer (ablation)DARE · BeerAdvocate Appearancetoken F1: 82.9 -> 41.3Hurt
Replace CLUB_NCE with CLUB as MI minimizer (ablation)DARE · BeerAdvocate Smelltoken F1: 61.2 -> 58.3Hurt
Replace CLUB_NCE with CLUB as MI minimizer (ablation)DARE · BeerAdvocate Palatetoken F1: 69.9 -> 67.4Hurt
Replace CLUB_NCE with L1Out as MI minimizer (ablation)DARE · CAIL2018 Terms of Penaltymacro-F1: 30.65 -> 3.72Hurt
Replace CLUB_NCE with CLUB as MI minimizer (ablation)DARE · CAIL2018 Chargesmacro-F1: 84.36 -> 74.84Hurt
Add reconstruction loss to DARE (ablation)DARE · BeerAdvocate Appearancetoken F1: 82.9 -> 82.6No clear effect
Replace RCNN with BigBird in DARE versus HardKumaDARE · BeerAdvocate Appearancetoken F1: 84.4 -> 87.5Improved
Use DARE instead of InfoCal(HK) with external guiderInfoCal(HK) · ERASER Movie Reviewstoken F1: 33.2 -> 33.4Improved
Use DARE instead of HardKumaHardKuma · Human evaluation, charge predictionusefulness: 3.95 -> 4.58Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy