AI Paper Atlas

Research map / Label-efficient and robust learning / Knowledge distillation

TrustAL: Trustworthy Active Learning Using Knowledge Distillation

Beong-woo Kwak, Youngwook Kim, Yu Jin Kim, Seung-won Hwang, Jinyoung Yeo

AAAI 2022 · Label-efficient and robust learning › Knowledge distillation · approach: Self and data-free distillation · 9 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
Use consistency-aware predecessor selection instead of the standalone acquisition modelTrustAL-NC with CONF · TRECcorrect consistency: 0.726 -> 0.774Improved
Use consistency-aware predecessor selection instead of the standalone acquisition modelTrustAL-NC with CONF · Movie reviewcorrect consistency: 0.637 -> 0.676Improved
Use consistency-aware predecessor selection instead of the standalone acquisition modelTrustAL-NC with CONF · SST-2correct consistency: 0.686 -> 0.730Improved
Choose the consistency-aware predecessor rather than the most recent predecessor as teacherTrustAL · TREC, Movie review, SST-2correct consistency: TREC: 0.743 -> 0.774; Movie review: 0.654 -> 0.676; SST-2: 0.705 -> 0.730Improved
Add predecessor-model distillation and consistency regularizationCONF, CORESET, and BADGE · TREC, Movie review, SST-2accuracyImproved
Use TrustAL-NC instead of a standalone acquisition baselineTrustAL with CONF, CORESET, or BADGE · TREClabeled data needed to reach baseline convergence accuracy: 40%; 30%; 50%Improved
Increase the distillation weight α from 0.75 to 10 (ablation)TrustAL-NC with CONF · TREC stable phaseaccuracy: 0.788({+0.011}) -> 0.810({+0.033})Improved
Increase the distillation weight α from 0.75 to 10 (ablation)TrustAL-NC with CONF · TREC stable phaseMCI: 17.94({-5.83}) -> 10.62({-13.15})Improved
Set distillation weight α to 20 rather than 10 (ablation)TrustAL-NC with CONF · TREC saturated phaseaccuracy: 0.892({+0.059}) -> 0.874({+0.041})Hurt
Increase the distillation weight α (ablation)TrustAL-NC · TRECaccuracy and MCIMixed
Use half-sized rather than the original development set (ablation)TrustAL-NC · TRECaccuracy: 0.008Hurt
Add consistency distillation when acquired labels are corruptedBADGE · TREC, after stable phase with 7% label noiseaccuracyImproved

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