Home▸AI/ML Drug Discovery Infrastructure▸2D REINVENT-Style Augmented-Likelihood Policy for Atom Chains: Prior, Target Hetero-Atom Fraction and the Gibbs Bound

🧪 2D REINVENT-Style Augmented-Likelihood Policy for Atom Chains: Prior, Target Hetero-Atom Fraction and the Gibbs Bound

Stacked policy bars for each chain position against the frozen prior, the latest generated atom chains with their scores, a REINVENT versus plain REINFORCE switch, and six independent numerical checks (exact enumeration, policy-gradient theorem, Gibbs identity).

AI/ML Drug Discovery Infrastructure2DAdvanced60 FPS📱 Mobile-adapted🟠 Blender⇄ 3D version
2D REINVENT-Style Augmented-Likelihood Policy for Atom Chains: Prior, Target Hetero-Atom Fraction and the Gibbs Bound ↗ Open standalone
⚙ Under the hood

Stacked policy bars for each chain position against the frozen prior, the latest generated atom chains with their scores, a REINVENT versus plain REINFORCE switch, and six independent numerical checks (exact enumeration, policy-gradient theorem, Gibbs identity).

REINVENTPolicy GradientReinforcement LearningGenerative ChemistryAugmented LikelihoodKL DivergenceGibbs Variational PrincipleAI Drug Discovery Infrastructure

2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install

What did you find?

Add reproduction steps (optional)