QM9
PulseAugur coverage of QM9 — every cluster mentioning QM9 across labs, papers, and developer communities, ranked by signal.
6 day(s) with sentiment data
-
New SIGMA objective improves molecular autoregressive models
Researchers have developed SIGMA, a novel objective for autoregressive molecular models that improves their ability to assign probabilities to molecules regardless of their serialization format. This method uses a dense…
-
New GNN models advance crystal property prediction and molecular simulations
Researchers have developed new graph neural network (GNN) models for predicting crystal properties. One approach, CPGN, uses a multi-scale GNN to jointly learn atomic, bond, and coordination-polyhedron representations, …
-
New method optimizes molecular graph generation via role-aware diffusion
Researchers have developed MotifRole-Diff, a novel approach to masked discrete diffusion for molecular graph generation. This method optimizes the corruption schedule by assigning different masking rates to molecular gr…
-
New SenCos-GEM framework enhances molecular property prediction accuracy
Researchers have developed SenCos-GEM, a new framework for molecular representation learning designed to improve the accuracy of predicting molecular properties. This approach integrates physics-guided geometric consist…
-
Decafs model improves generative AI interpretability and performance
Researchers have developed Decafs, a novel conditional generator based on Lie groups designed to improve the interpretability of flow-based generative models. By disentangling generative factors in the latent space thro…
-
New InertialAR model advances 3D molecule generation with Transformer++
Researchers have developed InertialAR, a novel autoregressive model for generating 3D molecules. This model addresses key challenges in molecule tokenization by creating an SE(3)- and permutation-invariant sequence of t…
-
New quantum and classical models achieve high accuracy in molecular property prediction with few parameters
Researchers have developed two new architectures, Iso-QGNN and Iso-CGNN, for molecular property prediction in low-data quantum chemistry settings. These models leverage a topology-aligned inductive bias, mirroring the m…
-
New MDM-VGB sampler enhances diffusion models with reward-guided remasking
Researchers have developed MDM-VGB, a novel discrete diffusion sampler designed to enhance Masked Diffusion Models (MDMs). This new method integrates reward-guided remasking, drawing inspiration from the Jerrum-Sinclair…
-
Molecular feature analysis challenges AI generalization heuristics
A new paper analyzes the spectral properties of molecular features to understand model generalization in machine learning. Researchers found that richer spectral features do not always lead to better performance, challe…
-
New VAE Enhances Molecular Generation by Improving Latent Space Smoothness
Researchers have developed TopVAE, a novel Variational Autoencoder (VAE) designed to improve the smoothness and validity of latent spaces in molecular diffusion models. Unlike previous methods relying on reconstruction …
-
Quantum-Inspired Methods Boost Machine Learning Representations
Researchers have developed new methods to enhance machine learning models by integrating quantum computing principles. One approach, QUIVER, uses quantum Fisher views to capture higher-order correlations in data, improv…
-
New tensor algebra embeds equivariance for symmetry discovery
Researchers have developed a new tensor algebra framework called $\star_G$ that intrinsically embeds equivariance, allowing for symmetry-preserving tensor approximation and physical symmetry discovery. This framework of…
-
Energy-based model generates physically consistent molecules
Researchers have developed EBMol, a novel energy-based model for generating physically consistent 3D molecules. This model learns an atom-additive potential without requiring explicit simulations during training, utiliz…
-
Orbital Transformers learn molecular wavefunctions for faster TDDFT simulations
Researchers have developed OrbEvo, an equivariant graph transformer model designed to predict molecular wavefunctions in time-dependent density functional theory (TDDFT). This new approach aims to accelerate the simulat…
-
NEAT transformer generates 3D molecules with state-of-the-art speed and accuracy
Researchers have developed NEAT, a novel autoregressive set transformer designed for 3D molecular generation. Unlike previous methods that rely on sequential atom ordering, NEAT treats molecules as sets and uses a neigh…
-
VQ-SAD model uses neuro-symbolic approach for improved molecule generation
Researchers have developed VQ-SAD, a novel neuro-symbolic model for molecule generation using diffusion techniques. This approach integrates symbolic information about atoms and bonds by treating them as latent variable…