Closed-Loop Transfer Enables AI to Yield Chemical Knowledge
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
AI-guided closed-loop experimentation has recently emerged as a promising method to optimize objective functions,1,2 but the substantial potential of this traditionally black-box approach to reveal new scientific knowledge has remained largely untapped. Here, we report a new AI-guided approach, dubbed Closed-Loop Transfer (CLT), that integrates closed-loop experiments with physics-based feature selection and supervised learning to yield new scientific knowledge in parallel with optimization of objective functions. CLT surprisingly revealed that high-energy regions of the triplet state manifold are paramount in dictating molecular photostability in solution across a diverse chemical library of light-harvesting donor-bridge-acceptor oligomers. Remarkably, this insight emerged after automated modular synthesis and experimental characterization of only ~1.5% of the theoretical chemical space. Supervised learning models considering millions of combinations of 100+ physics-based descriptors further showed that high energy triplet states most strongly correlate with photostability, while excluding more commonly considered predictors such as the lowest energy triplet state. The physics-informed model for photostability was even further confirmed and then strengthened using an explicit experimental test set, validating the substantial power of the CLT method. Broadly, these findings show that interfacing physics-based modeling with closed-loop discovery campaigns unimpeded by synthesis bottlenecks can rapidly illuminate fundamental chemical insights and guide more rational pursuit of frontier molecular functions.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.004 | 0.010 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it