Blog

This is a note from my experience in 2024. Please check the latest information from the official TACL website. The TACL submission process is a bit complex compared to recent conference submissions on OpenReview, but the editors (and editor assistants) are very helpful and responsive.

Influence function for neural networks is proposed in the ICML2017 best paper (Wei Koh & Liang, 2017). However, to the best of my knowledge, there is no generic PyTorch implementation with reliable test codes.

I have presented my research about implicit curve approximation in Meeting of the Minds 2019 (Undergraduate Research Symposium at CMU). My poster presented at Meeting of the Minds 2019 Implementation of present implicit curve approximation methods Our method is based on the topologically faithful implicit curve approximation method proposed by Keren (2004).

“Generating Sentences from a Continuous Space” (Bowman et al., 2015) https://github.com/ryokamoi/original_textvae Implementation of the first work on VAE for text. This model is a simple LSTM-LSTM seq2seq model with word dropout.