A latent mediation framework with sparse autoencoders enables non-additive token-level influence attribution in LLMs by learning orthogonal features and back-propagating attributions.
Disentangling by Factorising
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
abstract
We define and address the problem of unsupervised learning of disentangled representations on data generated from independent factors of variation. We propose FactorVAE, a method that disentangles by encouraging the distribution of representations to be factorial and hence independent across the dimensions. We show that it improves upon $\beta$-VAE by providing a better trade-off between disentanglement and reconstruction quality. Moreover, we highlight the problems of a commonly used disentanglement metric and introduce a new metric that does not suffer from them.
representative citing papers
DPA provides closed-form relation from level-set geometry to data score and proves extra latent components are conditionally independent, revealing intrinsic dimension.
A multi-branch β-VAE on tropical Pacific SST, OHC, and OLR fields yields a latent space that reconstructs data well and aligns with physical ENSO and longer-term coupled variability modes.
DAR replaces GAP with an attention-based aggregation module retrained jointly with the classifier head to disentangle core from spurious features and outperforms DFR on multiple datasets.
citing papers explorer
-
Correcting Influence: Unboxing LLM Outputs with Orthogonal Latent Spaces
A latent mediation framework with sparse autoencoders enables non-additive token-level influence attribution in LLMs by learning orthogonal features and back-propagating attributions.
-
Distributional Autoencoders Know the Score
DPA provides closed-form relation from level-set geometry to data score and proves extra latent components are conditionally independent, revealing intrinsic dimension.
-
What's in the latent space? Exploring coupled tropical Pacific variability within a Multi-branch $\beta$-Variational Autoencoder
A multi-branch β-VAE on tropical Pacific SST, OHC, and OLR fields yields a latent space that reconstructs data well and aligns with physical ENSO and longer-term coupled variability modes.
-
Deep Attention Reweighting: Post-Hoc Attention-Based Feature Aggregation in CNNs for Disentangling Core and Spurious Features under Spurious Correlations
DAR replaces GAP with an attention-based aggregation module retrained jointly with the classifier head to disentangle core from spurious features and outperforms DFR on multiple datasets.
- Learning to Theorize the World from Observation