Causal Intervention Framework for Variational Auto Encoder Mechanistic Interpretability
This provides tools for more transparent and controllable VAE architectures, addressing mechanistic interpretability for generative models, though it is incremental as it builds on existing VAE methods.
The paper tackled the challenge of understanding how Variational Autoencoders (VAEs) work by introducing a causal intervention framework to identify and analyze circuit motifs, showing that FactorVAE achieved higher disentanglement scores (0.084) and effect strengths (mean 4.59) compared to standard VAE (0.064, 3.99) and Beta-VAE (0.051, 3.43).
Mechanistic interpretability of deep learning models has emerged as a crucial research direction for understanding the functioning of neural networks. While significant progress has been made in interpreting discriminative models like transformers, understanding generative models such as Variational Autoencoders (VAEs) remains challenging. This paper introduces a comprehensive causal intervention framework for mechanistic interpretability of VAEs. We develop techniques to identify and analyze "circuit motifs" in VAEs, examining how semantic factors are encoded, processed, and disentangled through the network layers. Our approach uses targeted interventions at different levels: input manipulations, latent space perturbations, activation patching, and causal mediation analysis. We apply our framework to both synthetic datasets with known causal relationships and standard disentanglement benchmarks. Results show that our interventions can successfully isolate functional circuits, map computational graphs to causal graphs of semantic factors, and distinguish between polysemantic and monosemantic units. Furthermore, we introduce metrics for causal effect strength, intervention specificity, and circuit modularity that quantify the interpretability of VAE components. Experimental results demonstrate clear differences between VAE variants, with FactorVAE achieving higher disentanglement scores (0.084) and effect strengths (mean 4.59) compared to standard VAE (0.064, 3.99) and Beta-VAE (0.051, 3.43). Our framework advances the mechanistic understanding of generative models and provides tools for more transparent and controllable VAE architectures.