Introduces a causal intervention framework with new metrics for mechanistic interpretability of VAEs and reports empirical findings from extensive experiments on multiple models and datasets.
Testing relational understanding in text-guided image generation
3 Pith papers cite this work, alongside 39 external citations. Polarity classification is still indexing.
3
Pith papers citing it
39
external citations · external index
fields
cs.LG 3verdicts
UNVERDICTED 3representative citing papers
Compositionality emerges in neural networks only in a narrow depth-connectivity regime, with gradient descent converging to fractured solutions outside it.
Generative models learn rules before memorizing data, creating an innovation window whose width depends on dataset size and rule complexity, observed in both diffusion and autoregressive architectures.
citing papers explorer
-
A Multi-Level Causal Intervention Framework for Mechanistic Interpretability in Variational Autoencoders
Introduces a causal intervention framework with new metrics for mechanistic interpretability of VAEs and reports empirical findings from extensive experiments on multiple models and datasets.
-
Compositionality Emerges in a Narrow Depth-Connectivity Regime: Architecture Constraints and Solution Manifolds
Compositionality emerges in neural networks only in a narrow depth-connectivity regime, with gradient descent converging to fractured solutions outside it.
-
The two clocks and the innovation window: When and how generative models learn rules
Generative models learn rules before memorizing data, creating an innovation window whose width depends on dataset size and rule complexity, observed in both diffusion and autoregressive architectures.