Translation equivariant transformer neural processes

An implementation of the TE-TNP from the ICML 2024 paper "Translation Equivariant Transformer Neural Processes".

Dependencies

This code requires use of the cambridge-mlg/tnp library.

Citation

If you use this code for your research, please cite our paper:

@inproceedings{ashman2024translation,
	title        = {{Translation Equivariant Transformer Neural Processes}},
	author       = {Ashman, Matthew and Diaconu, Cristiana and Kim, Junhyuck and Sivaraya, Lakee and Markou, Stratis and Requeima, James and Bruinsma, Wessel P and Turner, Richard E},
	year         = 2024,
	booktitle    = {Forty-first International Conference on Machine Learning}
}