activity
20242026
collaborators

5 papers

cs.LG2026

SS-ESOAP: Self-Scaled Adaptive Preconditioning for Physics-Informed Learning

Guangyuan Wang, Mads Toftrup, Sebastian Loeschcke +2

Physics-informed neural networks (PINNs) often face ill-conditioned objectives that limit high-accuracy training. Dense quasi-Newton methods improve local conditioning but require…

cs.LG2026

M+Adam: Low-Precision Training via Additive-Multiplicative Optimization

Xiaoyuan Liang, Sebastian Loeschcke, Mads Toftrup +1

Training with quantized weights can reduce costs but often results in degraded accuracy, especially when optimization is carried out in low precision, without storing high-precisio…

cs.LG2025

TensorGRaD: Tensor Gradient Robust Decomposition for Memory-Efficient Neural Operator Training

Sebastian Loeschcke, David Pitt, Robert Joseph George +5

Scientific problems require resolving multi-scale phenomena across different resolutions and learning solution operators in infinite-dimensional function spaces. Neural operators p…

cs.CV2024

Coarse-To-Fine Tensor Trains for Compact Visual Representations

Sebastian Loeschcke, Dan Wang, Christian Leth-Espensen +3

The ability to learn compact, high-quality, and easy-to-optimize representations for visual data is paramount to many applications such as novel view synthesis and 3D reconstructio…

cs.LG2024

LoQT: Low-Rank Adapters for Quantized Pretraining

Sebastian Loeschcke, Mads Toftrup, Michael J. Kastoryano +2

Despite advances using low-rank adapters and quantization, pretraining of large models on consumer hardware has not been possible without model sharding, offloading during training…