Type Systems for Array and Tensor Shapes
Shapes of arrays and tensors, which are the number of dimensions and the size of each dimension, are error prone factors in programming. In particular, errors in dimension sizes are difficult to detect at compile time when they are parameterized in a program and transformed by library functions. We are interested in assisting programmers to avoid shape errors by designing and implementing type systems and tools that detect and alert errors.News
- Suda’s talk on Static Shape Checking for a Deep Learning Library at IPSJ SIG PRO Workshop
- Graduation March 2024
- Five Members Presented Bachelor’s Thesis