Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
The paper introduces the Instance-Adaptive Variational Autoencoder (IA-VAE), a novel approach to amortized variational inference that uses a hypernetwork to generate input-dependent modulations of a shared encoder. This allows for instance-specific adaptation of the inference model, reducing the amortization gap and improving posterior approximation accuracy. Experiments on synthetic and image datasets demonstrate that IA-VAE achieves performance comparable to standard encoders with fewer parameters and consistently improves held-out ELBO over baseline VAEs.
Instance-adaptive parameter modulation lets VAEs achieve better posterior approximations with fewer parameters, suggesting a more efficient use of model capacity.
Latent variable models, including variational autoencoders (VAE), remain a central tool in modern deep generative modeling due to their scalability and a well-founded probabilistic formulation. These models rely on amortized variational inference to enable efficient posterior approximation, but this efficiency comes at the cost of a shared parametrization, giving rise to the amortization gap. We propose the instance-adaptive variational autoencoder (IA-VAE), an amortized variational inference framework in which a hypernetwork generates input-dependent modulations of a shared encoder. This enables input-specific adaptation of the inference model while preserving the efficiency of a single forward pass. By leveraging instance-specific parameter modulations, the proposed approach can achieve performance comparable to standard encoders with substantially fewer parameters, indicating a more efficient use of model capacity. Experiments on synthetic data, where the true posterior is known, show that IA-VAE yields more accurate posterior approximations and reduces the amortization gap. Similarly, on standard image benchmarks, IA-VAE consistently improves held-out ELBO over baseline VAEs, with statistically significant gains across multiple runs. These results suggest that increasing the flexibility of the inference parametrization through instance-adaptive modulation is a key factor in mitigating amortization-induced suboptimality in deep generative models.