Language Models as AI Research World Models

Key points

• We investigate language models as Research World Models (RWMs), which predict the outcomes of candidate interventions across research environments.1
• Research knowledge acquired from real experimental experience improves RWM predictions of unseen interventions within the same environment (Spearman +0.27), and can be reused across environments.1
• Ablations across 13 language models used as RWMs show that adding research knowledge can improve intervention ranking more than changing models or increasing reasoning effort alone.1
• These findings support language models as RWMs and motivate accumulating experimental data for future RWM training.1

Sources (1)

• [1]Language Models as AI Research World Models
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 8, 04:18 PM

  AI research agents automate the cycle of proposing, implementing, and evaluating experiments, opening a path toward recursive self-improvement.

  We investigate language models as Research World Models (RWMs), which predict the outcomes of candidate interventions across research environments.

Extractive summary: sentences quoted from the sources.