Evaluating System-Level Reasoning in LLM Agents – Jacob Andreas

Jacob Andreas ask a question that sits at the heart of AI research: “How should machines learn?” Should they learn from datasets …

source