hypogenic - Automated Hypothesis Generation and Testing
Generates and tests hypotheses from observational datasets, optionally combining empirical patterns with research literature.
Tags
Updated: 2026-10-16Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Generate data-driven hypotheses
- Test competing hypotheses
- Integrate literature insights
- Refine hypothesis banks iteratively
- Process literature PDFs
- Run inference on hypotheses
Inputs
- Task configuration file
- Training, validation, and test datasets
- Research paper PDFs
- Hypothesis bank file
- LLM API credentials or local model
Outputs
- Generated hypothesis bank files
- Inference result files
- Processed literature data
Requirements
- Installed hypogenic package
- LLM API access or local LLM
- Required dataset JSON format
- GROBID for PDF preprocessing
- s2orc-doc2json for literature processing