Model-Free Risk-Sensitive Reinforcement Learning
arXiv:2111.02907
Abstract
We extend temporal-difference (TD) learning in order to obtain risk-sensitive, model-free reinforcement learning algorithms. This extension can be regarded as modification of the Rescorla-Wagner rule, where the (sigmoidal) stimulus is taken to be either the event of over- or underestimating the TD target. As a result, one obtains a stochastic approximation rule for estimating the free energy from i.i.d. samples generated by a Gaussian distribution with unknown mean and variance. Since the Gaussian free energy is known to be a certainty-equivalent sensitive to the mean and the variance, the learning rule has applications in risk-sensitive decision-making.
arXiv commentDeepMind Tech Report: 13 pages, 4 figures
References in corpus (8)
- Learning to reinforcement learn
- Path integrals and symmetry breaking for optimal control theory
- Optimal control as a graphical model inference problem
- AI Safety Gridworlds
- Exploration vs Exploitation vs Safety: Risk-averse Multi-Armed Bandits
- Meta-learning of Sequential Strategies
- If MaxEnt RL is the Answer, What is the Question?
- Meta-trained agents implement Bayes-optimal agents