About the
Psi Research Library
A curated, machine-readable archive of peer-reviewed research on psi phenomena, built for epistemic transparency and rigorous inquiry.
Our Mission
The Psi Research Library is a curated collection of peer-reviewed research on psi phenomena, built with epistemic transparency. Our goal is not to advocate for or against psi claims, but to present the evidence — both supportive and critical — in a structured, searchable format.
We serve researchers, students, curious skeptics, and AI agents alike by providing a neutral ground for examining the data. By standardizing metadata and methodology scores, we make it easier to compare studies across decades of research.
What Makes This Library Different
Most archives are passive. We actively structure data to highlight methodological quality and debate context.
Both Sides Included
Every major controversy includes papers arguing for and against the phenomenon. We do not hide negative results or replication failures.
Methodology Scores
Every paper is tagged with methodological metadata (sample size, controls, blinding) allowing you to filter by rigor, not just conclusion.
Skeptic Prominence
Critical papers and skeptical analyses are given equal prominence in search results and controversy maps to ensure balanced viewing.
🛡️ Core Principles
Epistemic Balance
We prioritize neutral categorization over advocacy. The library is a tool for inquiry, not a manifesto.
Methodological Transparency
Full documentation of how papers were selected, processed, and categorized, including known biases.
Open Access Data
The underlying catalog is available as JSON. We believe data about science should be open to all.
Machine Readable
Structured specifically for LLM consumption via llms.txt and API, enabling advanced AI analysis.
🏗️ How It Was Built
The library was curated using a hybrid approach: systematic literature review combined with AI-assisted processing.
Paper metadata was extracted and structured using large language models (Claude 3.5, Qwen 2.5), then rigorously validated against original sources by human curators. This allowed us to process a vast volume of literature while maintaining high accuracy in categorization and summarization.
- Sources: PubMed, PsycINFO, JPAR, Google Scholar
- Validation: Human-in-the-loop verification
- Bias Check: Automated flagging of potential conflicts
Use This Library
Reading Paths
Curated journeys for beginners and researchers to navigate the field.
JSON API
Access the full catalog programmatically for your own analysis.
llms.txt
Optimized context file for AI agents and LLM integration.
User Guide
How to search, filter, and interpret the data effectively.
Curated By
This project represents a collaboration between human critical thinking and machine efficiency to organize complex scientific data.