12 papers
Is Hierarchical Quantization Essential for Optimal Reconstruction?
Shirin Reyhanian, Laurenz Wiskott
Vector-quantized variational autoencoders (VQ-VAEs) are central to models that rely on high reconstruction fidelity, from neural compression to generative pipelines. Hierarchical e…
Probing Length Generalization in Mamba via Image Reconstruction
Jan Rathjens, Robin Schiewer, Laurenz Wiskott +1
Mamba has attracted widespread interest as a general-purpose sequence model due to its low computational complexity and competitive performance relative to transformers. However, i…
Effects of Distributional Biases on Gradient-Based Causal Discovery in the Bivariate Categorical Case
Tim Schwabe, Moritz Lange, Laurenz Wiskott +1
Gradient-based causal discovery shows great potential for deducing causal structure from data in an efficient and scalable way. Those approaches however can be susceptible to distr…
The Course Difficulty Analysis Cookbook
Frederik Baucks, Robin Schmucker, Laurenz Wiskott
Curriculum analytics (CA) studies curriculum structure and student data to ensure the quality of educational programs. An essential aspect is studying course properties, which invo…
Understanding Transformer-based Vision Models through Inversion
Jan Rathjens, Shirin Reyhanian, David Kappel +1
Understanding the mechanisms underlying deep neural networks remains a fundamental challenge in machine learning and computer vision. One promising, yet only preliminarily explored…
Object-centric Denoising Diffusion Models for Physical Reasoning
Moritz Lange, Raphael C. Engelhardt, Wolfgang Konen +2
Reasoning about the trajectories of multiple, interacting objects is integral to physical reasoning tasks in machine learning. This involves conditions imposed on the objects at di…