5 citations · 7 across the 6 of their papers we have counts for
7 papers
Local Sparsity Enables Unsupervised LLM Safety Detection
Xin Chen, Gil Kur, Alexander Shevchenko +1
Deployment-time safety methods for large language models (LLMs) are predominantly supervised and assume access to unsafe training data. Nevertheless, new attacks and harm categorie…
Fast Unlearning at Scale via Margin Self-Correction
Federico Di Gennaro, Alexander Shevchenko, Fanny Yang
Language-model unlearning updates a trained model to behave as if it had not seen selected training examples, while preserving utility and avoiding costly retraining. Existing appr…
Specialization after Generalization: Towards Understanding Test-Time Training in Foundation Models
Jonas Hübotter, Patrik Wolf, Alexander Shevchenko +3
Recent empirical studies have explored the idea of continuing to train a model at test-time for a given task, known as test-time training (TTT), and have found it to yield signific…
Compression of Structured Data with Autoencoders: Provable Benefit of Nonlinearities and Depth
Kevin Kögler, Alexander Shevchenko, Hamed Hassani +1
Autoencoders are a prominent model in many empirical branches of machine learning and lossy data compression. However, basic theoretical questions remain unanswered even in a shall…
Fundamental Limits of Two-layer Autoencoders, and Achieving Them with Gradient Methods
Alexander Shevchenko, Kevin Kögler, Hamed Hassani +1
Autoencoders are a popular model in many branches of machine learning and lossy data compression. However, their fundamental limits, the performance of gradient methods and the fea…
Mean-field Analysis of Piecewise Linear Solutions for Wide ReLU Networks
Alexander Shevchenko, Vyacheslav Kungurtsev, Marco Mondelli
Understanding the properties of neural networks trained via stochastic gradient descent (SGD) is at the heart of the theory of deep learning. In this work, we take a mean-field vie…
Landscape Connectivity and Dropout Stability of SGD Solutions for Over-parameterized Neural Networks
Alexander Shevchenko, Marco Mondelli
The optimization of multilayer neural networks typically leads to a solution with zero training error, yet the landscape can exhibit spurious local minima and the minima can be dis…