No "Zero-Shot" Without Exponential Data: Pretraining Concept Frequency Determines Multimodal Model Performance

Documenti analoghi

• Efficient Lifelong Model Evaluation in an Era of Rapid Progress
  di: Prabhu, Ameya, et al.
  Pubblicazione: (2024)
• ONEBench to Test Them All: Sample-Level Benchmarking Over Open-Ended Capabilities
  di: Ghosh, Adhiraj, et al.
  Pubblicazione: (2024)
• How to Merge Your Multimodal Models Over Time?
  di: Dziadzio, Sebastian, et al.
  Pubblicazione: (2024)
• A Sober Look at Progress in Language Model Reasoning: Pitfalls and Paths to Reproducibility
  di: Hochlehnert, Andreas, et al.
  Pubblicazione: (2025)
• A Practitioner's Guide to Continual Multimodal Pretraining
  di: Roth, Karsten, et al.
  Pubblicazione: (2024)