cd/entity/Negative Preference Optimizationยท homeโ€บ entitiesโ€บ Negative Preference Optimization
grep -l @negative preference optimization /news/*.json | wc -l โ†’ 1

Negative Preference Optimization

mentions 1 type Person feed RSS

// recent coverage 1 mentions

22:23
2026-07-06
aws.amazon.com
large-language-models

Teaching models to forget: Selective unlearning with Amazon Nova

Amazon Web Services introduced Reverse Direct Preference Optimization (rDPO), a novel unlearning technique behind Amazon Nova Customizable Content Moderation Settings (CCMS), which allows approved cusโ€ฆ

// co-occurs with top 6 entities