Goal-conditioned Reinforcement Learning (RL) aims at learning optimal policies, given goals encoded in special command inputs. Here we study goal-conditioned neural nets (NNs) that learn to generate deep NN policies in form of context-specific weight matrices, similar to Fast Weight Programmers and other methods from the 1990s. Using context commands of the form "generate a policy that achieves a desired expected return," our NN generators combine powerful exploration of parameter space with generalization across commands to iteratively find better and better policies. A form of weight-sharing HyperNetworks and policy embeddings scales our method to generate deep NNs. Experiments show how a single learned policy generator can produce policies that achieve any return seen during training. Finally, we evaluate our algorithm on a set of continuous control tasks where it exhibits competitive performance.
|Original language||English (US)|
|Title of host publication||ICML 2022 : 39th International Conference on Machine Learning|
|State||Published - 2022|
Bibliographical noteKAUST Repository Item: Exported on 2022-12-21
Acknowledgements: We thank Kazuki Irie, Mirek Strupl, Dylan Ashley, Róbert Csordás, Aleksandar Stanic and Anand ´ Gopalakrishnan for their feedback. This work was supported by the ERC Advanced Grant (no: 742870) and by the Swiss National Supercomputing Centre (CSCS, projects: s1090, s1154). We also thank NVIDIA Corporation for donating a DGX-1 as part of the Pioneers of AI Research Award and to IBM for donating a Minsky machine.