10 papers · 1 filter
TROLL: Trust Regions improve Reinforcement Learning for Large Language Models
Philipp Becker, Niklas Freymuth, Serge Thilges +2
Reinforcement Learning (RL) with PPO-like clip objectives has become the standard choice for reward-based fine-tuning of large language models (LLMs). Although recent work has expl…
Improving Long-Range Interactions in Graph Neural Simulators via Hamiltonian Dynamics
Tai Hoang, Alessandro Trenta, Alessio Gravina +4
Learning to simulate complex physical systems from data has emerged as a promising way to overcome the limitations of traditional numerical solvers, which often require prohibitive…
AMBER: Adaptive Mesh Generation by Iterative Mesh Resolution Prediction
Niklas Freymuth, Tobias Würth, Nicolas Schreiber +9
The cost and accuracy of simulating complex physical systems using the Finite Element Method (FEM) scales with the resolution of the underlying mesh. Adaptive meshes improve comput…
Geometry-aware RL for Manipulation of Varying Shapes and Deformable Objects
Tai Hoang, Huy Le, Philipp Becker +2
Manipulating objects with varying geometries and deformable objects is a major challenge in robotics. Tasks such as insertion with different objects or cloth hanging require precis…
Efficient Off-Policy Learning for High-Dimensional Action Spaces
Fabian Otto, Philipp Becker, Ngo Anh Vien +1
Existing off-policy reinforcement learning algorithms often rely on an explicit state-action-value function representation, which can be problematic in high-dimensional action spac…
Adaptive World Models: Learning Behaviors by Latent Imagination Under Non-Stationarity
Emiliyan Gospodinov, Vaisakh Shaj, Philipp Becker +2
Developing foundational world models is a key research direction for embodied intelligence, with the ability to adapt to non-stationary environments being a crucial criterion. In t…