Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2022
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2211.00086 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914627438247936 |
|---|---|
| author | Kooi, Jacob E. Hoogendoorn, Mark François-Lavet, Vincent |
| author_facet | Kooi, Jacob E. Hoogendoorn, Mark François-Lavet, Vincent |
| contents | In the context of MDPs with high-dimensional states, downstream tasks are predominantly applied on a compressed, low-dimensional representation of the original input space. A variety of learning objectives have therefore been used to attain useful representations. However, these representations usually lack interpretability of the different features. We present a novel approach that is able to disentangle latent features into a controllable and an uncontrollable partition. We illustrate that the resulting partitioned representations are easily interpretable on three types of environments and show that, in a distribution of procedurally generated maze environments, it is feasible to interpretably employ a planning algorithm in the isolated controllable latent partition. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2211_00086 |
| institution | arXiv |
| publishDate | 2022 |
| record_format | arxiv |
| spellingShingle | Disentangled (Un)Controllable Features Kooi, Jacob E. Hoogendoorn, Mark François-Lavet, Vincent Machine Learning Artificial Intelligence In the context of MDPs with high-dimensional states, downstream tasks are predominantly applied on a compressed, low-dimensional representation of the original input space. A variety of learning objectives have therefore been used to attain useful representations. However, these representations usually lack interpretability of the different features. We present a novel approach that is able to disentangle latent features into a controllable and an uncontrollable partition. We illustrate that the resulting partitioned representations are easily interpretable on three types of environments and show that, in a distribution of procedurally generated maze environments, it is feasible to interpretably employ a planning algorithm in the isolated controllable latent partition. |
| title | Disentangled (Un)Controllable Features |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2211.00086 |