Search Results for author:
Found 16 papers, 5 papers with code
Disentangling Polysemantic Channels in Convolutional Neural Networks
Mechanistic interpretability is concerned with analyzing individual components in a (convolutional) neural network (CNN) and how they form larger circuits representing decision mechanisms.
VITAL: More Understandable Feature Visualization through Distribution Alignment and Relevant Information Flow
Feature visualization (FV) is a powerful tool to decode what information neurons are responding to and hence to better understand the reasoning behind such networks.
Unlocking Open-Set Language Accessibility in Vision Models
We apply our method on 40 visual classifiers and demonstrate two primary applications: 1) building both label-free and zero-shot concept bottleneck models and therefore converting any classifier to be inherently-interpretable and 2) zero-shot decoding of visual features into natural language.
Now you see me! A framework for obtaining class-relevant saliency maps
Neural networks are part of daily-life decision-making, including in high-stakes settings where understanding and transparency are key.
Sailing in high-dimensional spaces: Low-dimensional embeddings through angle preservation
As such, LDEs have to be faithful to the original high-dimensional data, i. e., they should represent the relationships that are encoded in the data, both at a local as well as global scale.
Pruning neural network models for gene regulatory dynamics using data and domain knowledge
The practical utility of machine learning models in the sciences often hinges on their interpretability.
Finding Interpretable Class-Specific Patterns through Efficient Neural Search
Discovering patterns in data that best describe the differences between classes allows to hypothesize and reason about class-specific mechanisms.
Understanding and Mitigating Classification Errors Through Interpretable Token Patterns
Characterizing these errors in easily interpretable terms gives insight into whether a classifier is prone to making systematic errors, but also gives a way to act and improve the classifier.
Preserving local densities in low-dimensional embeddings
We show, however, that these methods fail to reconstruct local properties, such as relative differences in densities (Fig.
Plant 'n' Seek: Can You Find the Winning Ticket?
The lottery ticket hypothesis has sparked the rapid development of pruning algorithms that aim to reduce the computational costs associated with deep learning during training and model deployment.
Lottery Tickets with Nonzero Biases
The strong lottery ticket hypothesis holds the promise that pruning randomly initialized deep neural networks could offer a computationally efficient alternative to deep learning with stochastic gradient descent.
Label-Descriptive Patterns and Their Application to Characterizing Classification Errors
Characterizing these errors in easily interpretable terms gives insight into whether a classifier is prone to making systematic errors, but also gives a way to act and improve the classifier.
Federated Learning from Small Datasets
Federated learning allows multiple parties to collaboratively train a joint model without sharing local data.
Picking Daisies in Private: Federated Learning from Small Datasets
Federated learning allows multiple parties to collaboratively train a joint model without sharing local data.
Factoring out prior knowledge from low-dimensional embeddings
Low-dimensional embedding techniques such as tSNE and UMAP allow visualizing high-dimensional data and therewith facilitate the discovery of interesting structure.
What's in the Box? Exploring the Inner Life of Neural Networks with Robust Rules
In particular, we consider activation values of a network for given data, and propose to mine noise-robust rules of the form $X \rightarrow Y$ , where $X$ and $Y$ are sets of neurons in different layers.
