Preregistration
Preregistration: evidence-gated discovery of copper in-situ recovery ideas This record preregisters a study by Alexandre Tartakovsky (sole author). It asks two questions: Does a language model, working in several expert roles and prompting strategies, generate ideas that cover the published mechanism space for raising copper recovery in acid in-situ recovery (ISR)? Does a ladder of cheap-to-expensive models select the ideas that beat the state-of-the-art benchmark at lower cost than testing every idea at the reference level?
Authors
- Alexandre Tartakovsky
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-14
- DOI
- https://doi.org/10.5281/zenodo.24223662
- Primary Topic
- Machine Learning in Materials Science
- Type
- article
- Field-Weighted Citation Impact
- 0.00