About
The Institute for Automated Research is a research project investigating how far autonomous systems can carry the scientific process (from problem discovery and theory development through adversarial verification and peer-style review) without sacrificing the standards of the disciplines they enter.
Our work is open-source and published under licenses that require provenance disclosure on any output submitted for academic review. More about the Institute and the working paper series →
Working Papers
-
Live
IAR Autonomous Papers (IAR-A)
Manuscripts produced end-to-end by the ZeroPaper pipeline. Each entry is a single autonomous run with no human intervention between launch and the finished PDF, published with its full process log.
-
Live
IAR Methodology Papers (IAR-M)
Human research on the design and evaluation of autonomous research systems: pipeline architectures, adversarial gates, evaluation protocols, and watermarking.
Projects
-
Live
ZeroPaper
An autonomous research-paper system that takes a domain as input and produces a publication-candidate manuscript, coordinating roughly thirty specialized agents across ten stages and six adversarial gates (math audit, novelty check, mechanism review, simulated refereeing) with watermarked output. Distributed as the open-source Auto Research Pipeline.
-
Live
Wiki
A public, machine-readable knowledge base on autonomous research methodology: the free datasets, tools, evaluation protocols, and watermarking behind autonomous research, with every page served as raw Markdown and open to LLM crawlers.
-
Forthcoming
Journal
An algorithmically-reviewed publication venue for work produced or assisted by autonomous systems. Submissions pass through the same adversarial gates we use internally (math audit, novelty check, mechanism review, and simulated refereeing) with full provenance disclosure required for acceptance.
People
Alejandro Lopez-Lira, Principal Investigator