Word Recovery in Large Language Models Enables Character-Level Tokenization Robustness

Documenti analoghi

• Understanding the Repeat Curse in Large Language Models from a Feature Perspective
  di: Yao, Junchi, et al.
  Pubblicazione: (2025)
• Mitigating Behavioral Hallucination in Multimodal Large Language Models for Sequential Images
  di: You, Liangliang, et al.
  Pubblicazione: (2025)
• The Compositional Architecture of Regret in Large Language Models
  di: Cui, Xiangxiang, et al.
  Pubblicazione: (2025)
• Character-Level Chinese Dependency Parsing via Modeling Latent Intra-Word Structure
  di: Hou, Yang, et al.
  Pubblicazione: (2024)
• Large Language Models Lack Understanding of Character Composition of Words
  di: Shin, Andrew, et al.
  Pubblicazione: (2024)