Rati Devidze
(რატი დევიძე*)
I am a final-year (all but dissertation) PhD student at the Machine Teaching group of Max Planck Institute for Software Systems: MPI-SWS, in Saarbrücken, Germany. I am also a member of the Saarbrücken Graduate School of Computer Science.
Throughout my tenure in Computer Science, I've been focused on advancing Reinforcement Learning methodologies. My primary research interests lie in formalizing reward design frameworks within the Reinforcement Learning framework. My research is to design informative and explicable reward functions, that provide clear signals to guide agents toward their goals by incentivizing optimal actions and discouraging incorrect ones via negative rewards.
For more information about me, please find my [C.V.]
Campus E1.5, Room 336
Max Planck Institute for Software Systems
66123, Saarbrücken
news
Jan 19, 2024 | Our paper Ethics in Action: Training Reinforcement Learning Agent for Moral Decision-making In Text-based Adventure Games was accepted to AISTATS'24. |
---|---|
Dec 15, 2023 | Our paper Informativeness of Reward Functions in Reinforcement Learning was accepted to AAMAS'24. |
Sep 14, 2022 | Our paper Exploration-Guided Reward Shaping for Reinforcement Learning under Sparse Rewards was accepted to NeurIPS'22. |
Sep 28, 2021 | Our paper Explicable Reward Design for Reinforcement Learning Agents was accepted to NeurIPS'21. |