51 citations · 228 across the 23 of their papers we have counts for
9 papers · 1 filter
SurvICL: a Foundation Model for in-context Survival Analysis
Julie Alberge, Linus Bleistein, Judith Abécassis +3
Tabular foundation models (TFMs) now outperform gradient-boosted trees in classification and regression. Yet in survival analysis, which predicts when an event occurs from censored…
V-ECE: Estimating General Expected Calibration Errors
Eugène Berta, Sacha Braun, David Holzmüller +2
In probabilistic classification, calibration error (CE) measures the average divergence of predicted probabilities from , the true class distribution for…
Conditional Coverage Diagnostics for Conformal Prediction
Sacha Braun, David Holzmüller, Michael I. Jordan +1
Evaluating conditional coverage remains one of the most persistent challenges in assessing the reliability of predictive systems. Although conformal methods can give guarantees on…
Beyond ReLU: How Activations Affect Neural Kernels and Random Wide Networks
David Holzmüller, Max Schölpple
In recent years, the neural tangent kernel (NTK) and neural network Gaussian process kernel (NNGP) have given theoreticians tractable limiting cases of fully connected neural netwo…
Mind the spikes: Benign overfitting of kernels and neural networks in fixed dimension
Moritz Haas, David Holzmüller, Ulrike von Luxburg +1
The success of over-parameterized neural networks trained to near-zero training error has caused great interest in the phenomenon of benign overfitting, where estimators are statis…
Convergence Rates for Non-Log-Concave Sampling and Log-Partition Estimation
David Holzmüller, Francis Bach
Sampling from Gibbs distributions and computing their log-partition function are fundamental tasks in statistics, machine learning, and statistical physics. While efficient algorit…
A Framework and Benchmark for Deep Batch Active Learning for Regression
David Holzmüller, Viktor Zaverkin, Johannes Kästner +1
The acquisition of labels for supervised learning can be expensive. To improve the sample efficiency of neural network regression, we study active learning methods that adaptively…
On the Universality of the Double Descent Peak in Ridgeless Regression
David Holzmüller
We prove a non-asymptotic distribution-independent lower bound for the expected mean squared generalization error caused by label noise in ridgeless linear regression. Our lower bo…
Training Two-Layer ReLU Networks with Gradient Descent is Inconsistent
David Holzmüller, Ingo Steinwart
We prove that two-layer (Leaky)ReLU networks initialized by e.g. the widely used method proposed by He et al. (2015) and trained using gradient descent on a least-squares loss are…