AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

Collaborative Alignment of NLP Models

Fereshte Khani, Marco Tulio Ribeiro

NeurIPS 2023 · Large language models › Language, safety and interpretability · approach: Controlled text generation · 0 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 CoAlign data instead of AdaTest data when adding a conflicting conceptconcept data · QQP, Corig 'X = not antonym (X)'examples broken by new concept: 24/50 -> 7/50Improved
use CoAlign data instead of AdaTest data for the conflicting concept pairconcept data · QQP, Corig 'X = not antonym (X)'examples fixed by new concept: 2/50 -> 5/50Improved
use CoAlign data instead of AdaTest data for similar concept pairconcept data · QQP, Corig 'X = synonym (X)'examples broken by new concept: 18/50 -> 9/50Improved
use CoAlign data instead of AdaTest data for similar concept pairconcept data · QQP, Corig 'X = synonym (X)'examples fixed by new concept: 18/50 -> 20/50Improved
run CoAlign on top of AdaTest-debugged dataAdaTest-debugged model · QQP 6 conceptsbugs revealedImproved
use CoAlign instead of naive data collection from biased seedBERT-base-uncased · SB (full concept)accuracy: 80.7 -> 94.5Improved
use naive data collection from biased seed instead of base modelBERT-base-uncased · SB (full concept)accuracy: 82.6 -> 80.7Hurt
use naive data collection vs CoAlign on the biased distributionBERT-base-uncased · biased-SBaccuracy: 98.6 -> 94.9Hurt
use CoAlign instead of random selectionRoBERTa-Large · MNLI two conceptsoverall average accuracy: 89.6 -> 91.9Improved
use CoAlign instead of uncertainty samplingRoBERTa-Large · MNLI two conceptsoverall average accuracy: 91.6 -> 91.9No clear effect
use CoAlign instead of random selectionRoBERTa-Large · MNLI two conceptsworst-case accuracy: 87.7 -> 90.7Improved
use CoAlign instead of uncertainty samplingRoBERTa-Large · Amazon 4 categoriesworst-case accuracy: 85.9 -> 86.0No clear effect

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