Completed Theses
▼ 2026 (1)
From pixels to plans: Converting image observations of grid domains into plans
Louis Maiworm
Bachelor’s thesis - RWTH Aachen University
▼ 2025 (13)
Learning General Policies for Qualitative Numeric Planning Problems
Marleen König
Bachelor’s thesis - RWTH Aachen University
On Limits of GNNs for Planning in Pushworld
Yannik Hesse
Master’s thesis - RWTH Aachen University
Learning Graph-Based Symbolic State Representations for Planning from Images
Jakob Gebler
Master’s thesis - RWTH Aachen University
Novelty-based Tree-of-Thought Search for LLM Reasoning and Planning
Leon Hamm
Bachelor’s thesis - RWTH Aachen University
Learning Robot Motion Policies for Pushing Objects that Generalize
Ümit Diker
Master’s thesis - RWTH Aachen University
Learning Symbolic State Descriptions From Images Using Scene Graph Generation
Nirmal Maheshwari
Master’s thesis - RWTH Aachen University
Action Classification Through Dynamic Scene Graph Generation From ImageSequences
Paul Schulte
Bachelor’s thesis - RWTH Aachen University
Investigating Policy Gradient Methods for Rubik’s Cube
Allegra Nagel
Bachelor’s thesis - RWTH Aachen University
Learning Critical Regions via Feasibility Constraints in Task and Motion Planning
Baran Dello
Bachelor’s thesis - RWTH Aachen University
Learning Lifted STRIPS Models from Partial Graphs with Reduced Action Labels
Niklas Jansen
Master’s thesis - RWTH Aachen University
Solving ARC as a Generalized Planning Problem using ASP
Jan Dornhege
Bachelor’s thesis - RWTH Aachen University
On Learning Representations in Model-Based Reinforcement Learning
Yingliu Lu
Bachelor’s thesis - RWTH Aachen University
▼ 2024 (3)
Exploring the Limits of Relational Graph Neural Networks in PushWorld
Samuel Stante
Bachelor’s thesis - RWTH Aachen University
Towards Generalizable Automated Driving: A Hierarchical Reinforcement Learning Approach with Engineered Low-Level Policies in a Scenario-Based Learning Environment
Mahmoud Abdelhamid
Master’s thesis - RWTH Aachen University
Jointly Learning Skill Execution and Domain Grounding for Robot Task and Motion Planning
Daniel Swoboda
Master’s thesis - RWTH Aachen University