Evaluating Inquiry Learning in LLM-Based Agents: Hypothesizing, Reasoning, and Exploration
Scientific inquiry involves iterative cycles of exploration, hypothesis generation, and reasoning over observations to uncover relationships in complex systems. Large language models (LLMs) have demonstrated strong capabilities in several components of this process, including inductive reasoning, hypothesis generation, and data analysis. However, prior work typically evaluates these abilities in…
We haven't written up this one. arXiv cs.AI has the full story — the link below goes straight to it.
