Neural Functional Transformers
This work addresses the problem of handling high-dimensional weight-space objects for researchers in machine learning, representing an incremental advancement with a novel method for a known bottleneck.
The paper tackled the challenge of constructing expressive and efficient neural functional architectures for processing neural network weights by introducing neural functional Transformers (NFTs), which use attention mechanisms to create permutation equivariant layers, resulting in improved performance matching or exceeding prior methods and boosting INR classification accuracy by up to +17%.
The recent success of neural networks as implicit representation of data has driven growing interest in neural functionals: models that can process other neural networks as input by operating directly over their weight spaces. Nevertheless, constructing expressive and efficient neural functional architectures that can handle high-dimensional weight-space objects remains challenging. This paper uses the attention mechanism to define a novel set of permutation equivariant weight-space layers and composes them into deep equivariant models called neural functional Transformers (NFTs). NFTs respect weight-space permutation symmetries while incorporating the advantages of attention, which have exhibited remarkable success across multiple domains. In experiments processing the weights of feedforward MLPs and CNNs, we find that NFTs match or exceed the performance of prior weight-space methods. We also leverage NFTs to develop Inr2Array, a novel method for computing permutation invariant latent representations from the weights of implicit neural representations (INRs). Our proposed method improves INR classification accuracy by up to $+17\%$ over existing methods. We provide an implementation of our layers at https://github.com/AllanYangZhou/nfn.