Microsoft Research blog
Provably efficient reinforcement learning with rich observations
| Akshay Krishnamurthy
Reinforcement learning, a machine learning paradigm for sequential decision making, has stormed into the limelight, receiving tremendous attention from both researchers and practitioners. When combined with deep learning, reinforcement learning (RL) has produced impressive empirical results, but the successes to…
Introducing MASS – A pre-training method that outperforms BERT and GPT in sequence to sequence language generation tasks
| Xu Tan
Editor’s note: Since 2018, pre-training has without a doubt become one of the hottest research topics in Natural Language Processing (NLP). By leveraging generalized language models like the BERT, GPT and XLNet, great breakthroughs have been achieved in natural language…
Reliability in Reinforcement Learning
| Romain Laroche
Reinforcement Learning (RL), much like scaling a 3,000-foot rock face, is about learning to make sequential decisions. The list of potential RL applications is expansive, spanning robotics (drone control), dialogue systems (personal assistants, automated call centers), the game industry (non-player…
Efficient inference for dynamical models using variational autoencoders
| Neil Dalchau
Dynamical systems theory provides a mathematical framework for studying how complex systems evolve over time, such as the neurons in our brains, the global climate system, or engineered cells. But predicting how these systems will behave in the future or…