Saved in:
Bibliographic Details
Main Authors: Chandramouli, Suyog, Kachergis, George, Jagadish, Akshay
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2604.25521
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915963127988224
author Chandramouli, Suyog
Kachergis, George
Jagadish, Akshay
author_facet Chandramouli, Suyog
Kachergis, George
Jagadish, Akshay
contents Cognitive science often evaluates theories through narrow paradigms and local model comparisons, limiting the integration of evidence across tasks and realizations. We introduce an automated adversarial collaboration framework for adjudicating among competing theories even when the candidate models and experiments must be discovered during the adjudication process. The system combines LLM-based theory agents, program synthesis, and information-theoretic experimental design in a closed loop. In a simulation study spanning three classic categorization theories, the framework recovered the ground-truth theory across noise settings with weaker reliability in the hardest settings. Together, the framework and findings provide a concrete proof of concept for closed-loop, in-silico theory adjudication in cognitive science.
format Preprint
id arxiv_https___arxiv_org_abs_2604_25521
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Automated Adversarial Collaboration for Advancing Theory Building in the Cognitive Sciences
Chandramouli, Suyog
Kachergis, George
Jagadish, Akshay
Artificial Intelligence
Cognitive science often evaluates theories through narrow paradigms and local model comparisons, limiting the integration of evidence across tasks and realizations. We introduce an automated adversarial collaboration framework for adjudicating among competing theories even when the candidate models and experiments must be discovered during the adjudication process. The system combines LLM-based theory agents, program synthesis, and information-theoretic experimental design in a closed loop. In a simulation study spanning three classic categorization theories, the framework recovered the ground-truth theory across noise settings with weaker reliability in the hardest settings. Together, the framework and findings provide a concrete proof of concept for closed-loop, in-silico theory adjudication in cognitive science.
title Automated Adversarial Collaboration for Advancing Theory Building in the Cognitive Sciences
topic Artificial Intelligence
url https://arxiv.org/abs/2604.25521