- Improving the Interpretability of Deep Neural Networks with Knowledge Distillation - [Arxiv] [QA]
- Adaptive Confidence Smoothing for Generalized Zero-Shot Learning - [Arxiv] [QA]
- Grounded Human-Object Interaction Hotspots from Video - [Arxiv] [QA]
- Abstracting Causal Models - [Arxiv] [QA]
- Face Completion with Semantic Knowledge and Collaborative Adversarial Learning - [Arxiv] [QA]
- Generalized Zero- and Few-Shot Learning via Aligned Variational Autoencoders - [Arxiv] [QA]
- Playing Text-Adventure Games with Graph-Based Deep Reinforcement Learning - [Arxiv] [QA]
- Deep Inception Generative Network for Cognitive Image Inpainting - [Arxiv] [QA]
- Strike (with) a Pose: Neural Networks Are Easily Fooled by Strange Poses of Familiar Objects - [Arxiv] [QA]
- Coordinate-based Texture Inpainting for Pose-Guided Image Generation - [Arxiv] [QA]
- GAN Dissection: Visualizing and Understanding Generative Adversarial Networks - [Arxiv] [QA]
- Self-Supervised Video Representation Learning with Space-Time Cubic Puzzles - [Arxiv] [QA]
- Bytes are All You Need: End-to-End Multilingual Speech Recognition and Synthesis with Bytes - [Arxiv] [QA]
- Causal Inference by String Diagram Surgery - [Arxiv] [QA]
- Transferable Interactiveness Knowledge for Human-Object Interaction Detection - [Arxiv] [QA]
- Generalized Zero-Shot Recognition based on Visually Semantic Embedding - [Arxiv] [QA]
- Scalable agent alignment via reward modeling: a research direction - [Arxiv] [QA]
- On Hallucinating Context and Background Pixels from a Face Mask using Multi-scale GANs - [Arxiv] [QA]
- Grasp2Vec: Learning Object Representations from Self-Supervised Grasping - [Arxiv] [QA]
- Reward learning from human preferences and demonstrations in Atari - [Arxiv] [QA]
- CGMH: Constrained Sentence Generation by Metropolis-Hastings Sampling - [Arxiv] [QA]
- Model Evaluation, Model Selection, and Algorithm Selection in Machine Learning - [Arxiv] [QA]
- Generative Dual Adversarial Network for Generalized Zero-shot Learning - [Arxiv] [QA]
- Blockwise Parallel Decoding for Deep Autoregressive Models - [Arxiv] [QA]
- Image Chat: Engaging Grounded Conversations - [Arxiv] [QA]
- Persistent-Homology-based Machine Learning and its Applications -- A Survey - [Arxiv] [QA]
- Taking Human out of Learning Applications: A Survey on Automated Machine Learning - [Arxiv] [QA]
- Automatically Evolving CNN Architectures Based on Blocks - [Arxiv] [QA]
- Image Inpainting via Generative Multi-column Convolutional Neural Networks - [Arxiv] [QA]
- Graph HyperNetworks for Neural Architecture Search - [Arxiv] [QA]
- BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding - [Arxiv] [QA]
- Batch Active Preference-Based Learning of Reward Functions - [Arxiv] [QA]
- Robustness via Retrying: Closed-Loop Robotic Manipulation with Self-Supervised Learning - [Arxiv] [QA]
- Learning Long-Range Perception Using Self-Supervision from Short-Range Sensors and Odometry - [Arxiv] [QA]
- TVQA: Localized, Compositional Video Question Answering - [Arxiv] [QA]
- AISHELL-2: Transforming Mandarin ASR Research Into Industrial Scale - [Arxiv] [QA]
- iCAN: Instance-Centric Attention Network for Human-Object Interaction Detection - [Arxiv] [QA]
- Discriminative Deep Dyna-Q: Robust Planning for Dialogue Policy Learning - [Arxiv] [QA]
- All You Need is "Love": Evading Hate-speech Detection - [Arxiv] [QA]
- Learning Human-Object Interactions by Graph Parsing Neural Networks - [Arxiv] [QA]
- Everybody Dance Now - [Arxiv] [QA]
- Pervasive Attention: 2D Convolutional Neural Networks for Sequence-to-Sequence Prediction - [Arxiv] [QA]
- Self-Supervised Model Adaptation for Multimodal Semantic Segmentation - [Arxiv] [QA]
- Learning Actionable Representations from Visual Observations - [Arxiv] [QA]
- Efficient Progressive Neural Architecture Search - [Arxiv] [QA]
- Regional Multi-scale Approach for Visually Pleasing Explanations of Deep Neural Networks - [Arxiv] [QA]
- Improving Spatiotemporal Self-Supervision by Deep Reinforcement Learning - [Arxiv] [QA]
- Diverse feature visualizations reveal invariances in early layers of deep neural networks - [Arxiv] [QA]
- Cross-lingual Argumentation Mining: Machine Translation (and a bit of Projection) is All You Need! - [Arxiv] [QA]
- Explaining Image Classifiers by Counterfactual Generation - [Arxiv] [QA]
- TESSERACT: Eliminating Experimental Bias in Malware Classification across Space and Time - [Arxiv] [QA]
- Towards Automated Deep Learning: Efficient Joint Neural Architecture and Hyperparameter Search - [Arxiv] [QA]
- Visual Reinforcement Learning with Imagined Goals - [Arxiv] [QA]
- Representation Learning with Contrastive Predictive Coding - [Arxiv] [QA]
- Talk the Walk: Navigating New York City through Grounded Dialogue - [Arxiv] [QA]
- A Tutorial on Bayesian Optimization - [Arxiv] [QA]
- Seq2RDF: An end-to-end application for deriving Triples from Natural Language Text - [Arxiv] [QA]
- A Benchmark for Interpretability Methods in Deep Neural Networks - [Arxiv] [QA]
- This Looks Like That: Deep Learning for Interpretable Image Recognition - [Arxiv] [QA]
- Video Inpainting by Jointly Learning Temporal Structure and Spatial Details - [Arxiv] [QA]
- RISE: Randomized Input Sampling for Explanation of Black-box Models - [Arxiv] [QA]
- Self-Supervised Feature Learning by Learning to Spot Artifacts - [Arxiv] [QA]
- Polynomial Regression As an Alternative to Neural Nets - [Arxiv] [QA]
- Resource-Efficient Neural Architect - [Arxiv] [QA]
- Auto-Meta: Automated Gradient Based Meta Learner Search - [Arxiv] [QA]
- Free-Form Image Inpainting with Gated Convolution - [Arxiv] [QA]
- Dank Learning: Generating Memes Using Deep Neural Networks - [Arxiv] [QA]
- A Peek Into the Hidden Layers of a Convolutional Neural Network Through a Factorization Lens - [Arxiv] [QA]
- Soccer on Your Tabletop - [Arxiv] [QA]
- Explaining Explanations: An Overview of Interpretability of Machine Learning - [Arxiv] [QA]
- How Important Is a Neuron? - [Arxiv] [QA]
- Rethinking Knowledge Graph Propagation for Zero-Shot Learning - [Arxiv] [QA]
- Theory and Experiments on Vector Quantized Autoencoders - [Arxiv] [QA]
- AutoAugment: Learning Augmentation Policies from Data - [Arxiv] [QA]
- Semantic Network Interpretation - [Arxiv] [QA]
- Controlling Personality-Based Stylistic Variation with Neural Natural Language Generators - [Arxiv] [QA]
- Progressive Ensemble Networks for Zero-Shot Recognition - [Arxiv] [QA]
- Unsupervised Learning of Neural Networks to Explain Neural Networks - [Arxiv] [QA]
- DVAE#: Discrete Variational Autoencoders with Relaxed Boltzmann Priors - [Arxiv] [QA]
- A Theoretical Explanation for Perplexing Behaviors of Backpropagation-based Visualizations - [Arxiv] [QA]
- Fighting Fake News: Image Splice Detection via Learned Self-Consistency - [Arxiv] [QA]
- SPG-Net: Segmentation Prediction and Guidance Network for Image Inpainting - [Arxiv] [QA]
- Unsupervised Feature Learning via Non-Parametric Instance-level Discrimination - [Arxiv] [QA]
- AGI Safety Literature Review - [Arxiv] [QA]
- Learnable PINs: Cross-Modal Embeddings for Person Identity - [Arxiv] [QA]
- How convolutional neural network see the world - A survey of convolutional neural network visualization methods - [Arxiv] [QA]
- FaceShop: Deep Sketch-based Face Image Editing - [Arxiv] [QA]
- Subgoal Discovery for Hierarchical Dialogue Policy Learning - [Arxiv] [QA]
- Image Inpainting for Irregular Holes Using Partial Convolutions - [Arxiv] [QA]
- Audio-Visual Scene Analysis with Self-Supervised Multisensory Features - [Arxiv] [QA]
- Bimonoidal Structure of Probability Monads - [Arxiv] [QA]
- Deep Painterly Harmonization - [Arxiv] [QA]
- The Sound of Pixels - [Arxiv] [QA]
- Learning to Separate Object Sounds by Watching Unlabeled Video - [Arxiv] [QA]
- Self-supervised Learning of Geometrically Stable Features Through Probabilistic Introspection - [Arxiv] [QA]
- The simple essence of automatic differentiation - [Arxiv] [QA]
- Meta-Learning Update Rules for Unsupervised Representation Learning - [Arxiv] [QA]
- MemGEN: Memory is All You Need - [Arxiv] [QA]
- Structural inpainting - [Arxiv] [QA]
- Learning Synergies between Pushing and Grasping with Self-supervised Deep Reinforcement Learning - [Arxiv] [QA]
- BAGAN: Data Augmentation with Balancing GAN - [Arxiv] [QA]
- Zero-shot Recognition via Semantic Embeddings and Knowledge Graphs - [Arxiv] [QA]
- Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge - [Arxiv] [QA]
- Preserving Semantic Relations for Zero-Shot Learning - [Arxiv] [QA]
- Relational Reasoning for Markov Chains in a Probabilistic Guarded Lambda Calculus - [Arxiv] [QA]
- Machine Theory of Mind - [Arxiv] [QA]
- Machine Theory of Mind - [Arxiv] [QA]
- Diversity is All You Need: Learning Skills without a Reward Function - [Arxiv] [QA]
- Multimodal Explanations: Justifying Decisions and Pointing to the Evidence - [Arxiv] [QA]
- DVAE++: Discrete Variational Autoencoders with Overlapping Transformations - [Arxiv] [QA]
- On Characterizing the Capacity of Neural Networks using Algebraic Topology - [Arxiv] [QA]
- A Survey Of Methods For Explaining Black Box Models - [Arxiv] [QA]
- Learning Image Representations by Completing Damaged Jigsaw Puzzles - [Arxiv] [QA]
- The Matrix Calculus You Need For Deep Learning - [Arxiv] [QA]
- Singularities in Einstein-conformally coupled Higgs cosmological models - [Arxiv] [QA]
- Interpreting CNNs via Decision Trees - [Arxiv] [QA]
- Shift-Net: Image Inpainting via Deep Feature Rearrangement - [Arxiv] [QA]
- Understanding Deep Architectures by Visual Summaries - [Arxiv] [QA]
- Generative Image Inpainting with Contextual Attention - [Arxiv] [QA]
- Net2Vec: Quantifying and Explaining how Concepts are Encoded by Filters in Deep Neural Networks - [Arxiv] [QA]