← Search

Valentin Hartmann

2 accepted papers

2024

Neural Redshift: Random Networks are not Random Functions

CVPR 2024poster

Our understanding of the generalization capabilities of neural networks NNs is still incomplete. Prevailing explanations are based on implicit biases of gradient descent GD but they cannot account for the capabilities of models from gradientfree methods nor the simplicity bias recently observed in u…

Cited by 19SourcePDFScholar
2020

Self-Supervised Learning of Scene-Graph Representations for Robotic Sequential Manipulation Planning

CoRL 2020

We present a self-supervised representation learning approach for visual reasoning and integrate it into a nonlinear program formulation for motion optimization to tackle sequential manipulation tasks. Such problems have usually been addressed by combined task and motion planning approaches, for whi