Contrastive Perplexity for Controlled Generation: An Application in Detoxifying Large Language Models

Documenti analoghi

• Learning Private Representations through Entropy-based Adversarial Training
  di: Klein, Tassilo, et al.
  Pubblicazione: (2025)
• Large Language Models can be Strong Self-Detoxifiers
  di: Ko, Ching-Yun, et al.
  Pubblicazione: (2024)
• Perplexed by Perplexity: Perplexity-Based Data Pruning With Small Reference Models
  di: Ankner, Zachary, et al.
  Pubblicazione: (2024)
• Detoxifying Large Language Models via Knowledge Editing
  di: Wang, Mengru, et al.
  Pubblicazione: (2024)
• Momentum Point-Perplexity Mechanics in Large Language Models
  di: Tomaz, Lorenzo, et al.
  Pubblicazione: (2025)