FusionCLM: enhanced molecular property prediction via knowledge fusion of chemical language models
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
Chemical Language Models (CLMs) have demonstrated capabilities in extracting patterns and predicting from vast volume of the Simplified Molecular Input Line Entry System (SMILES), a notation used to represent molecular structures. Different CLMs, developed from various architectures, can provide unique insights into molecular properties. To harness the uniqueness of different CLMs, we propose FusionCLM, a novel stacking-ensemble learning algorithm that integrate the outputs of multiple CLMs into a unified framework. FusionCLM first generates SMILES embeddings, predictions, and losses from each CLM. Auxiliary models are trained on these first-level predictions and embeddings to estimate test losses during inference. The losses and predictions are then concatenated to create an integrated feature matrix, which trains second-level meta-models for final predictions. Empirical testing on five datasets demonstrates that FusionCLM have better performance than individual CLM at the first level and three advanced multimodal deep learning frameworks, showcasing FusionCLM’s potential in advancing molecular property prediction. FusionCLM uses the stacking-ensemble learning method that integrates unique representation learning from multiple CLMs, allowing a more comprehensive learning of molecular SMILES data. This results in providing more accurate molecular property prediction, which can help in facilitating early discovery and development of promising drug candidates. By evaluating and comparing its performance against individual CLMs and existing multimodal deep learning frameworks, FusionCLM demonstrates improvements in prediction accuracy, distinguishing itself from prior models in this domain.
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 | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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