Research Practices
These commitments describe how the lab intends to work. Where a practice is planned rather than already in routine use, it says so.
- Reproducibility
- We aim to document methods, datasets, code, and dependencies so others can understand and repeat our analyses.
- Human review
- AI-assisted summaries, code, and interpretations should be checked by a researcher before informing conclusions.
- Source verification
- We check citations against original sources and distinguish evidence from assumptions.
- Responsible data use
- We prioritize public or appropriately licensed datasets and protect confidential information.
- Transparent reporting
- We clearly label proposed work, ongoing studies, preprints, and peer-reviewed publications.
- Rigorous evaluation
- We define evaluation criteria, compare against suitable baselines, and report limitations and unsuccessful results.
- Open collaboration
- Where permissions allow, we aim to share research methods, code, and findings for feedback and further study.