SynthWorlds: Controlled Parallel Worlds for Disentangling Reasoning and Knowledge in Language Models

Documenti analoghi

• What Are the Odds? Language Models Are Capable of Probabilistic Reasoning
  di: Paruchuri, Akshay, et al.
  Pubblicazione: (2024)
• Language Models Still Struggle to Zero-shot Reason about Time Series
  di: Merrill, Mike A., et al.
  Pubblicazione: (2024)
• Substance over Style: Evaluating Proactive Conversational Coaching Agents
  di: Srinivas, Vidya, et al.
  Pubblicazione: (2025)
• InvThink: Premortem Reasoning for Safer Language Models
  di: Kim, Yubin, et al.
  Pubblicazione: (2025)
• Transforming Wearable Data into Personal Health Insights using Large Language Model Agents
  di: Merrill, Mike A., et al.
  Pubblicazione: (2024)