8 papers
KV Cache Compression Through the Lens of Transform Coding
Hannah Laus, Claudio Mayrink Verdun, Hao Wang +2
The key-value (KV) cache stores information from past tokens and is a major memory bottleneck in long-context inference. Existing quantization methods address this bottleneck by re…
Plug-in Losses for Evidential Deep Learning: A Simplified Framework for Uncertainty Estimation that Includes the Softmax Classifier
Berk Hayta, Hannah Laus, Simon Mittermaier +1
Real-world sensor-based learning systems require uncertainty estimation that is both reliable and computationally efficient. Evidential Deep Learning (EDL) provides single-pass unc…
On Trajectory-Based Stability Analysis for -bit Sigma-Delta Quantization and its Application to the Second-Order Case
Rohan Joy, Felix Krahmer, Alessandro Lupoli
A state-of-the-art strategy for digitally representing a bandlimited signal is quantization. quantization schemes choose a bit sequence representing the s…
Low-Bit Quantization of Bandlimited Graph Signals via Iterative Methods
Felix Krahmer, He Lyu, Rayan Saab +3
We study the quantization of real-valued bandlimited signals on graphs, focusing on low-bit representations. We propose iterative noise-shaping algorithms for quantization, includi…
Solving Inverse Problems with Deep Linear Neural Networks: Global Convergence Guarantees for Gradient Descent with Weight Decay
Hannah Laus, Suzanna Parkinson, Vasileios Charisopoulos +2
Machine learning methods are commonly used to solve inverse problems, wherein an unknown signal must be estimated from few indirect measurements generated via a known acquisition p…
Phasebook: A Survey of Selected Open Problems in Phase Retrieval
Marc Allain, Selin Aslan, Wim Coene +13
Phase retrieval is an inverse problem that, on one hand, is crucial in many applications across imaging and physics, and, on the other hand, leads to deep research questions in the…