Research

Papers, hypotheses and experiments on AI and behavior. Claims and measured results belong to their stated evidence.

Manager research — papers, hypotheses, experiments and measured results from the existing research collection.

  1. Good Agent Evals Measure Work Done, Not Answers GivenResearch
  2. Mindfulness Bench v0: every small LLM fails at non-attachmentResearch
  3. AI ResponsibilityResearch
  4. BEMO: Integrating Behavioral Biomarkers and Economic Mechanisms for AI AccountabilityResearch