ChemBERTa: Molecular Property Prediction via Transformers
A systematic evaluation of RoBERTa transformers pretrained on 77M PubChem SMILES for molecular property prediction …
Machine-Learning
A Connection Between Score Matching and Denoising Autoencoders
Theoretical paper proving the equivalence between training Denoising Autoencoders and performing Score Matching on a …
Building Normalizing Flows with Stochastic Interpolants
A continuous-time normalizing flow using stochastic interpolants and quadratic loss to bypass costly ODE …
Flow Matching for Generative Modeling
A simulation-free framework for training Continuous Normalizing Flows using Conditional Flow Matching and Optimal …
Neural Ordinary Differential Equations
Introduces ODE-Nets, a continuous-depth neural network model parameterized by ODEs, enabling constant memory …
DynamicFlow: Integrating Protein Dynamics into Drug Design
Flow matching model that co-generates ligands and flexible protein pockets, addressing rigid-receptor limitations in …
InstructMol: Multi-Modal Molecular Assistant
A multi-modal LLM aligning 2D molecular graphs with text via two-stage instruction tuning for drug discovery tasks.
InvMSAFold: Fast & Diverse Inverse Folding
A fast, diverse inverse folding method combining deep learning with Potts models to capture full sequence landscapes.
MERMaid: Multimodal Reaction Mining
Vision-language pipeline extracting chemical reaction data from PDF figures and tables into structured knowledge graphs …
OCSAug: Diffusion-Based Augmentation for Hand-Drawn OCSR
A diffusion-based data augmentation pipeline (OCSAug) using DDPM and RePaint to improve optical chemical structure …
STOUT V2.0: SMILES to IUPAC Name Conversion
A Transformer-based model for translating SMILES to IUPAC names, trained on ~1 billion molecules, achieving ~99% …
Handwritten Chemical Ring Recognition with NNs
A two-phase neural network approach for recognizing handwritten heterocyclic chemical rings with ~94% accuracy.