Corpus ID: 214802310

Weakly-Supervised Reinforcement Learning for Controllable Behavior

@article{Lee2020WeaklySupervisedRL,
  title={Weakly-Supervised Reinforcement Learning for Controllable Behavior},
  author={L. Lee and Benjamin Eysenbach and R. Salakhutdinov and Shixiang Gu and Chelsea Finn},
  journal={ArXiv},
  year={2020},
  volume={abs/2004.02860}
}
Reinforcement learning (RL) is a powerful framework for learning to take actions to solve tasks. However, in many settings, an agent must winnow down the inconceivably large space of all possible tasks to the single task that it is currently being asked to solve. Can we instead constrain the space of tasks to those that are semantically meaningful? In this work, we introduce a framework for using weak supervision to automatically disentangle this semantically meaningful subspace of tasks from… Expand
Batch Exploration With Examples for Scalable Robotic Reinforcement Learning
Policy Learning Using Weak Supervision
Weakly Supervised Reinforcement Learning for Autonomous Highway Driving via Virtual Safety Cages

References

SHOWING 1-10 OF 84 REFERENCES
Reinforcement Learning with Unsupervised Auxiliary Tasks
Unsupervised Visuomotor Control through Distributional Planning Networks
Deep Reinforcement Learning from Human Preferences
Variational Inverse Control with Events: A General Framework for Data-Driven Reward Definition
Dynamical Distance Learning for Unsupervised and Semi-Supervised Skill Discovery
Hindsight Experience Replay
Near-Optimal Representation Learning for Hierarchical Reinforcement Learning
...
1
2
3
4
5
...