Workshop
Tue Dec 14 03:00 AM -- 11:00 AM (PST)
Bayesian Deep Learning
To deploy deep learning in the wild responsibly, we must know when models are making unsubstantiated guesses. The field of Bayesian Deep Learning (BDL) has been a focal point in the ML community for the development of such tools. Big strides have been made in BDL in recent years, with the field making an impact outside of the ML community, in fields including astronomy, medical imaging, physical sciences, and many others. But the field of BDL itself is facing an evaluation crisis: most BDL papers evaluate uncertainty estimation quality of new methods on MNIST and CIFAR alone, ignoring needs of real world applications which use BDL. Therefore, apart from discussing latest advances in BDL methodologies, a particular focus of this year’s programme is on the reliability of BDL techniques in downstream tasks. This focus is reflected through invited talks from practitioners in other fields and by working together with the two NeurIPS challenges in BDL — the Approximate Inference in Bayesian Deep Learning Challenge and the Shifts Challenge on Robustness and Uncertainty under Real-World Distributional Shift — advertising work done in applications including autonomous driving, medical, space, and more. We hope that the mainstream BDL community will adopt real world benchmarks based on such applications, pushing the field forward beyond MNIST and CIFAR evaluations.
| Opening Remarks (Opening remarks (zoom)) | |
| Adaptive and Robust Learning with Bayes (Invited talk) | |
| A Bayesian Perspective on Meta-Learning (Invited talk) | |
| Shifts Challenge: Robustness and Uncertainty under Real-World Distributional Shift (Competition talk) | |
| Gaussian Dropout as an Information Bottleneck Layer (Contributed talk) | |
| Funnels: Exact Maximum Likelihood with Dimensionality Reduction (Contributed talk) | |
| Posters (gather town link to the right) and lunch break (Poster) | |
| Spacecraft Collision Avoidance with Bayesian Deep Learning (Invited talk) | |
| Inference & Sampling with Symmetries (Invited talk) | |
| Bayesian Neural Networks, Andversarial Attacks, and How the Amount of Samples Matters (Invited talk) | |
| Posters (gather town) (Poster) | |
| Quantified Uncertainty for Safe Operation of Particle Accelerators (Invited talk) | |
| Diversity is All You Need to Improve Bayesian Model Averaging (Contributed talk) | |
| Structure Stochastic Gradient MCMC: a hybrid VI and MCMC approach (Contributed talk) | |
| Evaluating Approximate Inference in Bayesian Deep Learning (Competition talk) | |
| An Automatic Finite-Data Robustness Metric for Bayes and Beyond: Can Dropping a Little Data Change Conclusions? (Invited talk) | |
| Closing remarks | |
| Social and Posters (gather town) (Poster) | |
| Adversarial Learning of a Variational Generative Model with Succinct Bottleneck Representation (Poster) | |
| Robust outlier detection by de-biasing VAE likelihoods (Poster) | |
| The Dynamics of Functional Diversity throughout Neural Network Training (Poster) | |
| Biases in variational Bayesian neural networks (Poster) | |
| Bayesian Inference in Augmented Bow Tie Networks (Poster) | |
| Fast Finite Width Neural Tangent Kernel (Poster) | |
| Reliable Uncertainty Quantification of Deep Learning Models for a Free Electron Laser Scientific Facility (Poster) | |
| Latent Goal Allocation for Multi-Agent Goal-Conditioned Self-Supervised Learning (Poster) | |
| Constraining cosmological parameters from N-body simulations with Bayesian Neural Networks (Poster) | |
| Model-embedding flows: Combining the inductive biases of model-free deep learning and explicit probabilistic modeling (Poster) | |
| Likelihood-free Density Ratio Acquisition Functions are not Equivalent to Expected Improvements (Poster) | |
| Object-Factored Models with Partially Observable State (Poster) | |
| On Efficient Uncertainty Estimation for Resource-Constrained Mobile Applications (Poster) | |
| Dropout and Ensemble Networks for Thermospheric Density Uncertainty Estimation (Poster) | |
| Benchmark for Out-of-Distribution Detection in Deep Reinforcement Learning (Poster) | |
| Mixture-of-experts VAEs can disregard unimodal variation in surjective multimodal data (Poster) | |
| Depth Uncertainty Networks for Active Learning (Poster) | |
| The Peril of Popular Deep Learning Uncertainty Estimation Methods (Poster) | |
| Dependence between Bayesian neural network units (Poster) | |
| Precision Agriculture Based on Bayesian Neural Network (Poster) | |
| Funnels: Exact maximum likelihood with dimensionality reduction (Poster) | |
| Exploring the Limits of Epistemic Uncertainty Quantification in Low-Shot Settings (Poster) | |
| Progress in Self-Certified Neural Networks (Poster) | |
| Multimodal Relational VAE (Poster) | |
| Certifiably Robust Variational Autoencoders (Poster) | |
| Deep Classifiers with Label Noise Modeling and Distance Awareness (Poster) | |
| Benchmarking Bayesian Deep Learning on Diabetic Retinopathy Detection Tasks (Poster) | |
| Stochastic Local Winner-Takes-All Networks Enable Profound Adversarial Robustness (Poster) | |
| Pathologies in Priors and Inference for Bayesian Transformers (Poster) | |
| Analytically Tractable Inference in Neural Networks - An Alternative to Backpropagation (Poster) | |
| Posterior Temperature Optimization in Variational Inference for Inverse Problems (Poster) | |
| Relaxed-Responsibility Hierarchical Discrete VAEs (Poster) | |
| Decomposing Representations for Deterministic Uncertainty Estimation (Poster) | |
| Laplace Approximation with Diagonalized Hessian for Over-parameterized Neural Networks (Poster) | |
| Mixtures of Laplace Approximations for Improved Post-Hoc Uncertainty in Deep Learning (Poster) | |
| Contrastive Generative Adversarial Network for Anomaly Detection (Poster) | |
| On Symmetries in Variational Bayesian Neural Nets (Poster) | |
| Greedy Bayesian Posterior Approximation with Deep Ensembles (Poster) | |
| On Feature Collapse and Deep Kernel Learning for Single Forward Pass Uncertainty (Poster) | |
| An Empirical Study of Neural Kernel Bandits (Poster) | |
| Structured Stochastic Gradient MCMC: a hybrid VI and MCMC approach (Poster) | |
| Contrastive Representation Learning with Trainable Augmentation Channel (Poster) | |
| Power-law asymptotics of the generalization error for GP regression under power-law priors and targets (Poster) | |
| Deep Bayesian Learning for Car Hacking Detection (Poster) | |
| Uncertainty Baselines: Benchmarks for Uncertainty & Robustness in Deep Learning (Poster) | |
| Uncertainty Quantification in End-to-End Implicit Neural Representations for Medical Imaging (Poster) | |
| Evaluating Predictive Uncertainty and Robustness to Distributional Shift Using Real World Data (Poster) | |
| Generalization Gap in Amortized Inference (Poster) | |
| Information-theoretic stochastic contrastive conditional GAN: InfoSCC-GAN (Poster) | |
| Being a Bit Frequentist Improves Bayesian Neural Networks (Poster) | |
| Resilience of Bayesian Layer-Wise Explanations under Adversarial Attacks (Poster) | |
| Reproducible, incremental representation learning with Rosetta VAE (Poster) | |
| Non-stationary Gaussian process discriminant analysis with variable selection for high-dimensional functional data (Poster) | |
| Generation of data on discontinuous manifolds via continuous stochastic non-invertible networks (Poster) | |
| An Empirical Comparison of GANs and Normalizing Flows for Density Estimation (Poster) | |
| Evaluating Deep Learning Uncertainty Quantification Methods for Neutrino Physics Applications (Poster) | |
| Unveiling Mode-connectivity of the ELBO Landscape (Poster) | |
| Can Network Flatness Explain the Training Speed-Generalisation Connection? (Poster) | |
| Reflected Hamiltonian Monte Carlo (Poster) | |
| Infinite-channel deep convolutional Stable neural networks (Poster) | |
| Revisiting the Structured Variational Autoencoder (Poster) | |
| Gaussian dropout as an information bottleneck layer (Poster) | |
| Kronecker-Factored Optimal Curvature (Poster) | |
| Reducing redundancy in Semantic-KITTI: Study on data augmentations within Active Learning (Poster) | |
| Diversity is All You Need to Improve Bayesian Model Averaging (Poster) | |
| Regularizations Are All You Need: Weather Prediction Under Distributional Shift (Poster) | |
| An Empirical Analysis of Uncertainty Estimation in Genomics Applications (Poster) | |
| Hierarchical Topic Evaluation: Statistical vs. Neural Models (Poster) | |
| Federated Functional Variational Inference (Poster) | |
| Towards Robust Object Detection: Bayesian RetinaNet for Homoscedastic Aleatoric Uncertainty Modeling (Poster) | |
| Stochastic Pruning: Fine-Tuning, and PAC-Bayes bound optimization (Poster) | |