Explanation through Reward Model Reconciliation using POMDP Tree Search
This addresses user trust in AI for mission-critical applications, but it is incremental as it builds on existing POMDP and reward modeling techniques.
The paper tackled the problem of reconciling differences between an AI algorithm's reward model and a human user's implicit reward model in POMDP planning, using action discrepancies to estimate user objectives as reward function weightings.
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.