ICLR 2018 accepted papers on Probabilistic and scientific modeling
34 of the 337 papers accepted at ICLR 2018 (10.1%) are on Probabilistic and scientific modeling, compared with 11.6% at ICLR 2017. They fall into 7 topics.
See ICLR 2018 in the interactive map
Submitting to ICLR? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Variational and generative sampling · 18 papers
- Wasserstein Auto-Encoders (517 citations)
- Learning Latent Permutations with Gumbel-Sinkhorn Networks (148 citations)
- AmbientGAN: Generative models from lossy measurements (102 citations)
- Backpropagation through the Void: Optimizing control variates for black-box gradient estimation (93 citations)
- On Unifying Deep Generative Models (91 citations)
- Large Scale Optimal Transport and Mapping Estimation (73 citations)
- Auto-Encoding Sequential Monte Carlo (71 citations)
- WHAI: Weibull Hybrid Autoencoding Inference for Deep Topic Modeling (63 citations)
- Gradient Estimators for Implicit Models (51 citations)
- Variational Inference of Disentangled Latent Concepts from Unlabeled Observations (50 citations)
- Generative Models of Visually Grounded Imagination (47 citations)
- Debiasing Evidence Approximations: On Importance-weighted Autoencoders and Jackknife Variational Inference (34 citations)
and 6 more in the interactive map.
Molecules and proteins · 6 papers
- Syntax-Directed Variational Autoencoder for Structured Data (232 citations)
- Twin Networks: Matching the Future for Sequence Generation (57 citations)
- Learning Approximate Inference Networks for Structured Prediction (44 citations)
- Learning a Generative Model for Validity in Complex Discrete Structures (11 citations)
- Automatically Inferring Data Quality for Spatiotemporal Forecasting (4 citations)
- Learning Sparse Latent Representations with the Deep Copula Information Bottleneck
Dynamical systems · 4 papers
- Emergence of grid-like representations by training recurrent neural networks to perform spatial localization (203 citations)
- Simulating Action Dynamics with Neural Process Networks (26 citations)
- Synthesizing realistic neural population activity patterns using Generative Adversarial Networks (25 citations)
- Learning Deep Mean Field Games for Modeling Large Population Behavior (8 citations)
Uncertainty and Gaussian processes · 3 papers
- Deep Neural Networks as Gaussian Processes (553 citations)
- Recasting Gradient-Based Meta-Learning as Hierarchical Bayes (155 citations)
- Gaussian Process Behaviour in Wide Deep Neural Networks (89 citations)
Differential equations and PDEs · 1 paper
Treatment effects · 1 paper
Causal discovery · 1 paper
- Implicit Causal Models for Genome-wide Association Studies (16 citations)
