Newest papers

Post-Hoc Feature Selection Layer for Neural Networks Interpretability

João Kenji Suwa, João Pedro Silveira e Silva, Bruno Iochins Grisci · 2026 · arXiv

DOI: 10.48550/arXiv.2610.31420 · open-access PDF

Indonesian

Interpretabilitas jaringan saraf kompleks tetap menjadi tantangan penting, terutama untuk model yang sudah diimplementasikan di domain taruhan tinggi. Untuk mengatasi hal ini, kami menyesuaikan Layer Seleksi Fitur (FSL) untuk mengoptimalkan beban fitur setelah pelatihan, varian yang kita sebut post-hoc FSL. Pendekatan kami mengubah FSL sebagai modul yang ringan dan dapat dilatih yang mengintegrasikan dengan model yang sudah dibekukan sebelumnya di dataset tabel untuk menyoroti fitur yang dianggap paling penting oleh model asli.

Machine translation (Meta M2M100); the English original follows.

Abstract

The interpretability of complex neural networks remains a critical challenge, especially for models already deployed in high-stakes domains. To address this, we adapted the Feature Selection Layer (FSL) to optimize feature weights post-training, a variant we term post-hoc FSL. Our approach reframes the FSL as a lightweight, trainable module that integrates with already frozen pre-trained models on tabular datasets to highlight the features the original model considers most important. This post-hoc FSL learns feature relevance by fine-tuning its weights based on the original model's learned out

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange