hypogenic - Generate and test scientific hypotheses
Generates and tests scientific hypotheses from observational data, literature insights, or their combination using large language models.
Tags
Updated: 2026-10-12Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Generate data-driven hypotheses
- Integrate literature insights
- Refine hypothesis banks
- Test competing hypotheses
- Process research PDFs
- Run parallel inference
Inputs
- Task configuration file
- Training dataset
- Validation dataset
- Test dataset
- Hypothesis bank file
- Research paper PDFs
- Generation method
- Model selection
- Hypothesis count
Outputs
- Generated hypothesis bank
- Inference results
- Processed literature files
- Cached LLM responses
Requirements
- Python package installation
- Large language model access
- HuggingFace-format datasets
- YAML task configuration
- Redis server for caching
- s2orc-doc2json for literature PDFs
- GROBID for PDF preprocessing