Explanation through Reward Model Reconciliation using POMDP Tree Search
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2023
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910160713154560 |
|---|---|
| author | Kraske, Benjamin D. Saksena, Anshu Buczak, Anna L. Sunberg, Zachary N. |
| author_facet | Kraske, Benjamin D. Saksena, Anshu Buczak, Anna L. Sunberg, Zachary N. |
| contents | As artificial intelligence (AI) algorithms are increasingly used in mission-critical applications, promoting user-trust of these systems will be essential to their success. Ensuring users understand the models over which algorithms reason promotes user trust. This work seeks to reconcile differences between the reward model that an algorithm uses for online partially observable Markov decision (POMDP) planning and the implicit reward model assumed by a human user. Action discrepancies, differences in decisions made by an algorithm and user, are leveraged to estimate a user's objectives as expressed in weightings of a reward function. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2305_00931 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Explanation through Reward Model Reconciliation using POMDP Tree Search Kraske, Benjamin D. Saksena, Anshu Buczak, Anna L. Sunberg, Zachary N. Artificial Intelligence Human-Computer Interaction Machine Learning As artificial intelligence (AI) algorithms are increasingly used in mission-critical applications, promoting user-trust of these systems will be essential to their success. Ensuring users understand the models over which algorithms reason promotes user trust. This work seeks to reconcile differences between the reward model that an algorithm uses for online partially observable Markov decision (POMDP) planning and the implicit reward model assumed by a human user. Action discrepancies, differences in decisions made by an algorithm and user, are leveraged to estimate a user's objectives as expressed in weightings of a reward function. |
| title | Explanation through Reward Model Reconciliation using POMDP Tree Search |
| topic | Artificial Intelligence Human-Computer Interaction Machine Learning |
| url | https://arxiv.org/abs/2305.00931 |