Born a Transformer -- Always a Transformer? On the Effect of Pretraining on Architectural Abilities

Documenti analoghi

• The Expressive Capacity of State Space Models: A Formal Language Perspective
  di: Sarrof, Yash, et al.
  Pubblicazione: (2024)
• Contextualize-then-Aggregate: Circuits for In-Context Learning in Gemma-2 2B
  di: Bakalova, Aleksandra, et al.
  Pubblicazione: (2025)
• On the Ability of Transformers to Verify Plans
  di: Sarrof, Yash, et al.
  Pubblicazione: (2026)
• Barriers to Universal Reasoning With Transformers (And How to Overcome Them)
  di: Kraus, Oliver, et al.
  Pubblicazione: (2026)
• Can LLMs subtract numbers?
  di: Jobanputra, Mayank, et al.
  Pubblicazione: (2025)