Pre-Trained Language Models Represent Some Geographic Populations Better Than Others

Documenti analoghi

• FS-RAG: A Frame Semantics Based Approach for Improved Factual Accuracy in Large Language Models
  di: Madabushi, Harish Tayyar
  Pubblicazione: (2024)
• Scaling Policy Compliance Assessment in Language Models with Policy Reasoning Traces
  di: Imperial, Joseph Marvin, et al.
  Pubblicazione: (2025)
• Code-Mixed Probes Show How Pre-Trained Models Generalise On Code-Switched Text
  di: De Leon, Frances A. Laureano, et al.
  Pubblicazione: (2024)
• Neither Stochastic Parroting nor AGI: LLMs Solve Tasks through Context-Directed Extrapolation from Training Data Priors
  di: Madabushi, Harish Tayyar, et al.
  Pubblicazione: (2025)
• Standardize: Aligning Language Models with Expert-Defined Standards for Content Generation
  di: Imperial, Joseph Marvin, et al.
  Pubblicazione: (2024)