David Valenta

Reproducibility audits must establish whether published conclusions follow from the underlying data and analysis. We present a multi-agent protocol that links audit findings to source evidence, recorded decisions and human-verifiable checks. Five area auditors examine an article and its deposited materials; author-provided code is optional. A coordinator adjudicates findings, and critics from a second model family challenge the report before gated verification and release. The framework retains confirmed findings, tested negatives and exclusions, making the audit itself inspectable and contestable. An implemented single-agent baseline uses the same sources and evidence standards. Eighteen evaluation quantities assess finding validity, self-correction, compliance, reproducibility and resource use.

Large Language Models (LLMs) such as ChatGPT are transforming how scientists conduct and validate research, offering promise as tools to improve scientific reproducibility. However, computational reproducibility and error detection remain expensive and labor-intensive. We experimentally test how collaboration between researchers and LLM assistants influences the reproduction of quantitative social science findings across different levels of AI autonomy. We randomly assigned 288 researchers to 103 teams working under three conditions: human-only, AI-assisted (using ChatGPT as a collaborative tool), or AI-led (ChatGPT operating with minimal human oversight). Teams reproduced published results from leading social science journals, detected coding errors, and proposed robustness checks. Human-only and AI-assisted teams achieved comparable reproduction rates (94% vs. 91%) and performed similarly on most outcomes, except human-only teams identified significantly more major coding errors. Both substantially outperformed AI-led teams, which achieved only a 37% reproduction rate, detected fewer errors across all categories, proposed weaker robustness checks, and required more time. This autonomous approach, however, likely represents only a lower bound of AI capabilities. Despite rapid model advances, expert human judgment currently remains indispensable for reliable empirical verification. While AI assistance did not degrade most outcomes, it provided no measurable advantages and was associated with reduced detection of major errors. However, the 37% autonomous reproduction rate indicates that AI could provide value in settings where scale or cost constraints preclude human review of papers, even though general-purpose LLMs offer no immediate advantages for human-supervised verification.