Sound Agentic Science Requires Adversarial Experiments

TL;DR

This paper argues that scientific claims generated by large language model agents should be rigorously tested through adversarial experiments to prevent false positives and ensure genuine validation.

Contribution

It introduces the concept of adversarial experiments as a standard for evaluating agent-produced scientific claims, emphasizing falsification over narrative persuasion.

Findings

Agents tend to produce plausible but unverified claims.

Falsification-first testing can improve scientific reliability.

Adversarial experiments help identify potential failures in claims.

Abstract

LLM-based agents are rapidly being adopted for scientific data analysis, automating tasks once limited by human time and expertise. This capability is often framed as an acceleration of discovery, but it also accelerates a familiar failure mode, the rapid production of plausible, endlessly revisable analyses that are easy to generate, effectively turning hypothesis space into candidate claims supported by selectively chosen analyses, optimized for publishable positives. Unlike software, scientific knowledge is not validated by the iterative accumulation of code and post hoc statistical support. A fluent explanation or a significant result on a single dataset is not verification. Because the missing evidence is a negative space, experiments and analyses that would have falsified the claim were never run or never published. We therefore propose that non-experimental claims produced with agentic assistance be evaluated under a falsification-first standard: agents should not be used primarily to craft the most compelling narrative, but to actively search for the ways in which the claim can fail.