no code implementations • 11 Apr 2024 • Simon Schrodi, David T. Hoffmann, Max Argus, Volker Fischer, Thomas Brox
This revealed that the driving factor behind both, the modality gap and the object bias, is the information imbalance between images and captions.
1 code implementation • 5 Feb 2024 • Riccardo Grazzi, Julien Siems, Simon Schrodi, Thomas Brox, Frank Hutter
State of the art foundation models such as GPT-4 perform surprisingly well at in-context learning (ICL), a variant of meta-learning concerning the learned ability to solve tasks during a neural network forward pass, exploiting contextual information provided as input to the model.
no code implementations • 19 Oct 2023 • David T. Hoffmann, Simon Schrodi, Nadine Behrmann, Volker Fischer, Thomas Brox
In this work, we study rapid, step-wise improvements of the loss in transformers when being confronted with multi-step decision tasks.
no code implementations • 10 Oct 2023 • Karim Farid, Simon Schrodi, Max Argus, Thomas Brox
LDCE harnesses the capabilities of recent class- or text-conditional foundation latent diffusion models to expedite counterfactual generation and focus on the important, semantic parts of the data.
1 code implementation • 9 Oct 2023 • Simon Schrodi, Ferdinand Briegel, Max Argus, Andreas Christen, Thomas Brox
We show the efficacy of our approach across a wide spectrum of study areas and time scales.
2 code implementations • NeurIPS 2023 • Simon Schrodi, Danny Stoll, Binxin Ru, Rhea Sukthanker, Thomas Brox, Frank Hutter
In this work, we introduce a unifying search space design framework based on context-free grammars that can naturally and compactly generate expressive hierarchical search spaces that are 100s of orders of magnitude larger than common spaces from the literature.
1 code implementation • ICML Workshop AutoML 2021 • Julia Guerrero-Viu, Sven Hauns, Sergio Izquierdo, Guilherme Miotto, Simon Schrodi, Andre Biedenkapp, Thomas Elsken, Difan Deng, Marius Lindauer, Frank Hutter
Neural architecture search (NAS) and hyperparameter optimization (HPO) make deep learning accessible to non-experts by automatically finding the architecture of the deep neural network to use and tuning the hyperparameters of the used training pipeline.
1 code implementation • CVPR 2022 • Simon Schrodi, Tonmoy Saikia, Thomas Brox
We show how these mistakes can be rectified in order to make optical flow networks robust to physical patch-based attacks.