Top 10 Papers From ICLR 2020 Conference

• ALBERT: A Lite BERT for Self-supervised Learning of Language Representations
• Plug and Play Language Models: A Simple Approach to Controlled Text Generation
• Meta-Learning without Memorization
• Reformer: The Efficient Transformer
• Rapid Learning or Feature Reuse? Towards Understanding the Effectiveness of MAML
• Your Classifier is Secretly an Energy Based Model and You Should Treat it Like One
• Playing the lottery with rewards and multiple languages: lottery tickets in RL and NLP
• On Identifiability in Transformers
• Neural Tangents: Fast and Easy Infinite Neural Networks in Python
• Generalization through Memorization: Nearest Neighbor Language Models