7 papers
Optimistic Task Inference for Behavior Foundation Models
Thomas Rupf, Marco Bagatella, Marin Vlastelica +1
Behavior Foundation Models (BFMs) are capable of retrieving high-performing policy for any reward function specified directly at test-time, commonly referred to as zero-shot reinfo…
Flow Density Control: Generative Optimization Beyond Entropy-Regularized Fine-Tuning
Riccardo De Santi, Marin Vlastelica, Ya-Ping Hsieh +3
Adapting large-scale foundation flow and diffusion generative models to optimize task-specific objectives while preserving prior information is crucial for real-world applications…
Epistemically-guided forward-backward exploration
Núria Armengol UrpÃ, Marin Vlastelica, Georg Martius +1
Zero-shot reinforcement learning is necessary for extracting optimal policies in absence of concrete rewards for fast adaptation to future problem settings. Forward-backward repres…
Provable Maximum Entropy Manifold Exploration via Diffusion Models
Riccardo De Santi, Marin Vlastelica, Ya-Ping Hsieh +3
Exploration is critical for solving real-world decision-making problems such as scientific discovery, where the objective is to generate truly novel designs rather than mimic exist…
Dual-Force: Enhanced Offline Diversity Maximization under Imitation Constraints
Pavel Kolev, Marin Vlastelica, Georg Martius
Offline diversity maximization under imitation constraints can transform demonstration data into a set of distinct behavioral policies, improving robustness to distribution shift w…
Causal Action Influence Aware Counterfactual Data Augmentation
Núria Armengol UrpÃ, Marco Bagatella, Marin Vlastelica +1
Offline data are both valuable and practical resources for teaching robots complex behaviors. Ideally, learning agents should not be constrained by the scarcity of available demons…