Search on the Replay Buffer: Bridging Planning and Reinforcement Learning

NeurIPS 2019 google-research/google-research

Our algorithm, search on the replay buffer (SoRB), enables agents to solve sparse reward tasks over hundreds of steps, and generalizes substantially better than standard RL algorithms.

Optimizing Generalized Rate Metrics with Three Players

NeurIPS 2019 google-research/google-research

We present a general framework for solving a large class of learning problems with non-linear functions of classification rates.

Memory Efficient Adaptive Optimization

NeurIPS 2019 google-research/google-research

Adaptive gradient-based optimizers such as Adagrad and Adam are crucial for achieving state-of-the-art performance in machine translation and language modeling.

LANGUAGE MODELLING MACHINE TRANSLATION

A Benchmark for Interpretability Methods in Deep Neural Networks

NeurIPS 2019 google-research/google-research

We propose an empirical measure of the approximate accuracy of feature importance estimates in deep neural networks.

FEATURE IMPORTANCE IMAGE CLASSIFICATION

Differentiable Ranking and Sorting using Optimal Transport

NeurIPS 2019 google-research/google-research

From this observation, we propose extended rank and sort operators by considering optimal transport (OT) problems (the natural relaxation for assignments) where the auxiliary measure can be any weighted measure supported on $m$ increasing values, where $m \ne n$.

Reducing the variance in online optimization by transporting past gradients

NeurIPS 2019 google-research/google-research

While variance reduction methods have shown that reusing past gradients can be beneficial when there is a finite number of datapoints, they do not easily extend to the online setting.

STOCHASTIC OPTIMIZATION

Practical and Consistent Estimation of f-Divergences

NeurIPS 2019 google-research/google-research

The estimation of an f-divergence between two probability distributions based on samples is a fundamental problem in statistics and machine learning.

REPRESENTATION LEARNING

Unsupervised learning of object structure and dynamics from videos

NeurIPS 2019 google-research/google-research

Extracting and predicting object structure and dynamics from videos without supervision is a major challenge in machine learning.

CONTINUOUS CONTROL OBJECT TRACKING VIDEO PREDICTION

Stand-Alone Self-Attention in Vision Models

NeurIPS 2019 google-research/google-research

The natural question that arises is whether attention can be a stand-alone primitive for vision models instead of serving as just an augmentation on top of convolutions.

OBJECT DETECTION

Similarity of Neural Network Representations Revisited

ICML 2019 google-research/google-research

We introduce a similarity index that measures the relationship between representational similarity matrices and does not suffer from this limitation.