Tasks SotA Datasets Papers Methods Submit About

Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable Benchmarks All SotA Datasets Papers Methods

Community

Submit Results About

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Methods/Shifted Softplus

Shifted Softplus

GeneralIntroduced 200020 papers

Description

Shifted Softplus is an activation function ${\rm ssp}(x) = \ln( 0.5 e^{x} + 0.5 )$ , which SchNet employs as non-linearity throughout the network in order to obtain a smooth potential energy surface. The shifting ensures that ${\rm ssp}(0) = 0$ and improves the convergence of the network. This activation function shows similarity to ELUs, while having infinite order of continuity.

Papers Using This Method

Uncertainty Quantification in Graph Neural Networks with Shallow Ensembles2025-04-17 Machine learning surrogate models of many-body dispersion interactions in polymer melts2025-03-19 Machine Learned Force Fields: Fundamentals, its reach, and challenges2025-03-07 OpenQDC: Open Quantum Data Commons2024-11-29 SynCoTrain: A Dual Classifier PU-learning Framework for Synthesizability Prediction2024-11-18 Distribution Learning for Molecular Regression2024-07-30 Lightweight Geometric Deep Learning for Molecular Modelling in Catalyst Discovery2024-04-05 On the importance of catalyst-adsorbate 3D interactions for relaxed energy predictions2023-10-10 End-to-end AI framework for interpretable prediction of molecular and crystal properties2022-12-21 Machine Learning for Screening Large Organic Molecules2022-11-23 Developing Machine-Learned Potentials for Coarse-Grained Molecular Simulations: Challenges and Pitfalls2022-09-26 Predicting Aqueous Solubility of Organic Molecules Using Deep Learning Models with Varied Molecular Representations2021-05-26 The Open Catalyst 2020 (OC20) Dataset and Community Challenges2020-10-20 Heterogeneous Molecular Graph Neural Networks for Predicting Molecule Properties2020-09-26 A deep neural network for molecular wave functions in quasi-atomic minimal basis representation2020-05-11 Combining SchNet and SHARC: The SchNarc machine learning approach for excited-state dynamics2020-02-17 Learning representations of molecules and materials with atomistic neural networks2018-12-11 Analysis of Atomistic Representations Using Weighted Skip-Connections2018-10-23 Neural Message Passing with Edge Updates for Predicting Properties of Molecules and Materials2018-06-08 SchNet: A continuous-filter convolutional neural network for modeling quantum interactions2017-06-26