Augmented Hill-Climb
A reinforcement-learning rule that trains a molecule generator on the top scorers per batch, so costly scores like docking are used less.
The loop
- Proposes
SMILES strings from an RNN (also tested on a transformer) molecule generator
- Judged by
A scoring function evaluated on each sampled batch; the paper's tasks include docking objectives, with diversity filters
- Steers
RL updates the generator with a hybrid of REINVENT's loss and Hill-Climb's top-scorer selection
Headline result
On the authors' benchmark tasks, compared with REINVENT, optimization ability improved about 1.5-fold and sample efficiency about 45-fold, and Augmented Hill-Climb outperformed the other reinforcement learning strategies tested on six tasks.
What this does not prove
The gains are measured on the authors' benchmark tasks and docking objectives, not on lab-measured activity. It is a training rule inside the REINVENT-style setup rather than a separate system.
How it generalizes the harness
The generator is a SMILES RNN or transformer whose weights are updated by the score, with docking-style scoring as the judge.
Classification
| What it proposes | Molecule |
|---|---|
| Generative model | SMILES RNN (and RL-stabilized transformer), updated by reinforcement learning |
| Algorithmic authority | Task scoring functions including docking, plus diversity filters |
| Feedback | Reward signal |
| What gets updated | Model weights |
| When it runs | During training |
| Validity | Approximate physics (simulation, not lab results) |
| Evidence | Experiments reported |
| Lineage | Derivative |