Explainability in RL

Agent Model-Explaining

• [paper - reduction-of-imitation-learning-and)
• [paper
• [paper - 2303/PiRL)
• [paper - Decision-Tree)
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper - agent-neurosym-transformers)
• [paper
• [paper - symbolic-optimization)
• [paper - symbolic-optimization)
• [paper
• [paper
• [paper
• [paper
• [paper - learning)
• [paper - state-explanations)
• [paper - group/xai-nav-under-uncertainty-neurips2021)
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper - k/RRL)
• [paper - /Conservative-Q-Improvement)
• [paper
• [paper
• [paper
• [paper - Causal-Reinforcement-Learning)
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper - symbolic-optimization)
• [paper
• [paper
• [paper
• [paper - 2303/PiRL)
• [paper - reduction-of-imitation-learning-and)
• [paper
• [paper - kobayashi/cocoa)
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper

Reward Explaining

• [paper - multi-agent-policy-gradients#code)
• [paper
• [paper - global-reward-game-and-credit)
• [paper - PRL)
• [paper
• [paper - ILIAD/ELLA)
• [paper
• [paper - 2021-Repository)
• [paper
• [paper
• [paper - sg/PatchAIL)

State Explaining

• [paper
• [paper
• [paper
• [paper
• [paper
• [paper - Wrapper-Network-ICLR23)
• [paper - goel/MOREL)
• [paper - and-understanding-atari-agents)
• [paper
• [paper - DQN)
• [paper - KG)
• [paper
• [paper
• [paper
• [paper
• [paper - peng/HEX-RL)
• [paper - reinforcement-learning-lab/SVERL_icml_2023)
• [paper
• [paper - did-you-think-would-happen-explaining#code)
• [paper
• [paper - DQN)
• [paper - randomized-input-sampling-for)
• [paper - attention/)
• [paper - to-Interpret-Atari-Agents)
• [paper
• [paper - couthouis/xai-in-rl)
• [paper - elements-for-explainable)
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper - of-self-interpretable-agents)
• [paper
• [paper
• [paper - and-understanding-atari-agents)
• [paper - umass/saliency_maps)
• [paper
• [paper
• [paper

Task Explaining

• [paper
• [paper
• [paper - as-an-abstraction-for-hierarchical#code)
• [paper - task-reinforcement-learning-with#code)
• [paper
• [paper
• [paper
• [paper - task-reinforcement-learning-with#code)

🏠 Explainable AI Library

Subtask Scheduling with Human Annotation

• Aequitas
• Alibi Explain
• Captum
• DeepVis Toolbox - visualization-toolbox) |
• ELI5 - Memex/eli5) |
• IBM AI Explainability 360 - AI/AIX360) |
• iModels
• LIME
• OmniXAI
• SHAP

🤖️ Human knowledge for RL paradigm

Dense Reward on Human Command

• [paper - learn)

Fuzzy Controller Representing Human Knowledge

• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper
• [paper