A compact record of the research themes behind the current quantitative portfolio: financial prediction, structured and sequence representations, and conditional biomedical generation.
This repository provides context, not additional trading strategies or source releases.
| Theme | Question | Public boundary |
|---|---|---|
| Financial prediction | How should structure, uncertainty, and model risk enter return prediction? | The carbon-risk work is a paper retrospective because the original code is unavailable. |
| Structured and sequence models | When can explicit relationships and efficient long-sequence representations add useful information? | HUGS is presented as coauthored research with no module-level personal claim. |
| Biomedical generation | How can conditional generation and screening workflows be reasoned about safely? | Team-research context only; no component ownership, candidate count, or private experimental detail. |
- Publications — peer-reviewed and preprint records.
- Research themes — the intellectual connection to quantitative research.
- Contribution boundaries — explicit limits on personal and technical claims.
Implementation details, third-party source, and private research artifacts are not distributed publicly.