paper-with-me

홈 › Papers

On the limits of neural network explainability via descrambling

2023-01-18 · Shashank Sule, Richard G. Spencer, Wojciech Czaja

We characterize the exact solutions to neural network descrambling--a mathematical model for explaining the fully connected layers of trained neural networks (NNs). By reformulating the problem to the minimization of the Brockett function arising in graph matching and complexity theory we show that the principal components of the hidden layer preactivations can be characterized as the optimal explainers or descramblers for the layer weights, leading to descrambled weight matrices. We show that in typical deep learning contexts these descramblers take diverse and interesting forms including (1) matching largest principal components with the lowest frequency modes of the Fourier basis for isotropic hidden data, (2) discovering the semantic development in two-layer linear NNs for signal recovery problems, and (3) explaining CNNs by optimally permuting the neurons. Our numerical experiments indicate that the eigendecompositions of the hidden layer data--now understood as the descramblers--can also reveal the layer's underlying transformation. These results illustrate that the SVD is more directly related to the explainability of NNs than previously thought and offers a promising avenue for discovering interpretable motifs for the hidden action of NNs, especially in contexts of operator learning or physics-informed NNs, where the input/output data has limited human readability.

📄 PDF Abstract BibTeX arXiv:2301.07820

Code (1)

shashanksule/esvd 공식 구현

Tasks

Graph MatchingOperator learning

Similar Papers 제목 키워드 기반

The Limits of AI Explainability: An Algorithmic Information Theory Approach

2025-04-29 · Shrisha Rao

This paper establishes a theoretical foundation for understanding the fundamental limits of AI explainability through algorithmic information theory. We formalize explainability as the approximation of complex models by …

Automated Theorem Proving

Towards Multi-Grained Explainability for Graph Neural Networks

2021-12-01 · NeurIPS 2021 12 · Xiang Wang, Yingxin Wu, An Zhang, Xiangnan He 외

When a graph neural network (GNN) made a prediction, one raises question about explainability: “Which fraction of the input graph is most influential to the model’s decision?” Producing an answer requires understanding th…

Graph Neural Network

Improving Explainability of Sentence-level Metrics via Edit-level Attribution for Grammatical Error Correction

2024-12-17 · Takumi Goto, Justin Vasselli, Taro Watanabe

Various evaluation metrics have been proposed for Grammatical Error Correction (GEC), but many, particularly reference-free metrics, lack explainability. This lack of explainability hinders researchers from analyzing the…

AttributeGrammatical Error CorrectionSentence

SMILE: Self-Explainable Multimodal Information Bottleneck for Medical Diagnosis

2026-09-04 · Yuqing Yang, Alexander Schmatz, Zhaozhao Ma, Changkyu Choi 외 arxiv

Explainability is increasingly seen as a crucial requirement in AI-based medical diagnosis, particularly in safety-critical clinical decision-making. Most existing explainability methods in healthcare operate in a post-h…

Medical Diagnosis

Enhancing Visual Interpretability and Explainability in Functional Survival Trees and Forests

2025-04-25 · Giuseppe Loffredo, Elvira Romano, Fabrizio Maturo

Functional survival models are key tools for analyzing time-to-event data with complex predictors, such as functional or high-dimensional inputs. Despite their predictive strength, these models often lack interpretabilit…

Decision Making