… counterfactual explanations in supervised learning. Additionally, we explore the diferences
between counterfactual explanations in supervised learning … in reinforcement learning. Finally…

… learning models (eg differentiable or tree-based models). In this work, we propose a deep
reinforcement learning … us to generate batches of counterfactual instances in a single forward …

… explanations to an AI agent's behavior [Miller [38]]. In this work, we focus on generating
counterfactual explanations for deep reinforcement learning (… counterfactual state explanations, a …

… The latter aims at generating ex post explanations for … post-hoc explanations called
counterfactual explanations, which … from the predictive model, ie, counterfactual examples (CFs). …

… RELAX – the counterfactual explanation technique based on a deep reinforcement learning
… Our method is modelagnostic: it can generate valid explanations for any model that are …

… counterfactuals to be explained; something that is known to be a powerful explanation mech…
Our model generates counterfactual explanations by comparing causal chains of the actual …

… whether counterfactual explanations … counterfactual explanations for RL agents by
formulating the problem as a domain transfer problem which allows the use of adversarial learning …

… explanations of the behavior of deep reinforcement learning (… indicates that the derived
explanations are often unfalsifiable … approach grounded in counterfactual reasoning to test the …

… In this section, we describe RACCER, our approach for generating counterfactual explanations
for RL tasks. To generate a counterfactual explanation x′, we require oracle access to …

… Counterfactual explanations are a form of contrastive explanations. A counterfactual
explanation provides an answer to the question “Why p rather than q?”, where p is the fact which oc…