Learning for Intelligent Robotic Agents

Welcome to LiRA, where we focus on developing embodied AI agents. Unlike traditional passive AI systems, we specialize in crafting active AI agents capable of understanding and dynamically interacting with the physical world, much like humans do. Our goal extends to the seamless coexistence and collaboration between humans and robots, envisioning a future where these entities work harmoniously together. Drawing from an interdisciplinary toolkit, our research leverages methodologies from reinforcement learning, machine learning, deep learning, optimization, and probabilistic inference.

news

Jul 21, 2026	LiRA Lab participated in the 7th Turkish Robotics Conference held at Koç University, Istanbul, on July 19–21, 2026. Ece Kunduracioglu and Huzaifa Salahuddin delivered workshop presentations on behalf of Asst. Prof. Özgür S. Öğüz on the first day, while Asst. Prof. Öğüz gave a talk on the final day. Several lab members also presented their research during the poster sessions.
Jul 17, 2026	Our paper, “CoRAL: Contact-Rich Adaptive LLM-based Control for Robotic Manipulation”, by Berk Çiçek, Mert Kaan Er, and Özgür S. Öğüz, was published in the proceedings of Robotics: Science and Systems (RSS 2026).
Jul 11, 2026	Our paper, “Lagrangian Perturbation Diffusion Steering: Latent Reinforcement Learning for Generative Policies”, by Hikmet Şimşir and Özgür S. Öğüz, was published in the proceedings of the 43rd International Conference on Machine Learning (ICML 2026).
Jun 30, 2026	Our paper, “Uncertainty-Aware Safety Propagation Critics for Safe Reinforcement Learning”, by Kutay Demiray, Efe Eren Ceyani, and Özgür S. Öğüz, was published in Transactions on Machine Learning Research (TMLR) in June 2026.

selected publications

Uncertainty-Aware Safety Propagation Critics for Safe Reinforcement Learning

Kutay Demiray , Efe Eren Ceyani , and Ozgur S Oguz

Transactions on Machine Learning Research, 2026

Bib

@article{26-demiray2026-TMLR,
  title = {Uncertainty-Aware Safety Propagation Critics for Safe Reinforcement Learning},
  author = {Demiray, Kutay and Ceyani, Efe Eren and Oguz, Ozgur S},
  journal = {Transactions on Machine Learning Research},
  year = {2026},
}

CoRAL: Contact-Rich Adaptive LLM-based Control for Robotic Manipulation

Berk Cicek , Mert K Er , and Ozgur S Oguz

RSS 2026 Robotics: Science and Systems 2026, 2026

Bib

@article{25-tuncer2025-Arxiv,
  title = {CoRAL: Contact-Rich Adaptive LLM-based Control for Robotic Manipulation},
  author = {Cicek, Berk and Er, Mert K and Oguz, Ozgur S},
  journal = {RSS 2026 Robotics: Science and Systems 2026},
  year = {2026},
}

Lagrangian Perturbation Diffusion Steering: Latent Reinforcement Learning for Generative Policies

Hikmet Simsir , and Ozgur S Oguz

ICML 2026 Forty-Third International Conference on Machine Learning, 2026

Bib

@article{25-tuncer2025-Arxiw,
  title = {Lagrangian Perturbation Diffusion Steering: Latent Reinforcement Learning for Generative Policies},
  author = {Simsir, Hikmet and Oguz, Ozgur S},
  journal = {ICML 2026 Forty-Third International Conference on Machine Learning},
  year = {2026},
}

MO-SeGMan: Rearrangement Planning Framework for Multi Objective Sequential and Guided Manipulation in Constrained Environments

Cankut Bora Tuncer , Marc Toussaint , and Ozgur S Oguz

arXiv preprint arXiv:2511.01476, 2025

Bib

@article{25-tuncer2025-Arxix,
  title = {MO-SeGMan: Rearrangement Planning Framework for Multi Objective Sequential and Guided Manipulation in Constrained Environments},
  author = {Tuncer, Cankut Bora and Toussaint, Marc and Oguz, Ozgur S},
  journal = {arXiv preprint arXiv:2511.01476},
  year = {2025},
}

SeGMan: Sequential and Guided Manipulation Planner for Robust Planning in 2D Constrained Environments

Cankut Bora Tuncer , Dilruba Sultan Haliloglu , and Ozgur S Oguz

2025

Bib

@article{25-tuncer-IROS,
  title = {SeGMan: Sequential and Guided Manipulation Planner for Robust Planning in 2D Constrained Environments},
  author = {Tuncer, Cankut Bora and Haliloglu, Dilruba Sultan and Oguz, Ozgur S},
  booktitle = {2025 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  year = {2025},
  organization = {IEEE},
}

Way-Tu: A Framework for Tool Selection and Manipulation Using Waypoint Representations

Ece Kunduracioglu , and Ozgur S Oguz

In Dexterous Manipulation: Learning and Control with Diverse Modalities , 2025

Bib

@inproceedings{25-kunduracioglu-CoRLWorkshop,
  title = {Way-Tu: A Framework for Tool Selection and Manipulation Using Waypoint Representations},
  year = {2025},
  author = {Kunduracioglu, Ece and Oguz, Ozgur S},
  booktitle = {Dexterous Manipulation: Learning and Control with Diverse Modalities},
}

H-MaP: An Iterative and Hybrid Sequential Manipulation Planner

Berk Cicek , Arda Sarp Yenicesu , Cankut Bora Tuncer , and 2 more authors

IEEE Robotics and Automation Letters, 2025

Bib

@article{25-cicek-RA-L,
  title = {H-MaP: An Iterative and Hybrid Sequential Manipulation Planner},
  author = {Cicek, Berk and Yenicesu, Arda Sarp and Tuncer, Cankut Bora and Demiray, Kutay and Oguz, Ozgur S},
  journal = {IEEE Robotics and Automation Letters},
  year = {2025},
  publisher = {IEEE},
}

CUER: Corrected Uniform Experience Replay for Off-Policy Continuous Deep Reinforcement Learning Algorithms

Arda Sarp Yenicesu , Furkan B. Mutlu , Suleyman S. Kozat , and 1 more author

2024

arXiv Bib

@misc{24-yenicesu-Arxiv,
  title = {CUER: Corrected Uniform Experience Replay for Off-Policy Continuous Deep Reinforcement Learning Algorithms},
  author = {Yenicesu, Arda Sarp and Mutlu, Furkan B. and Kozat, Suleyman S. and Oguz, Ozgur S.},
  year = {2024},
}

Bilkent University
Department of Computer Engineering
Cankaya 06800 Ankara