Narutatsu Ri

Hi, my name is Edward, and I'm a rising senior at Columbia University. I'm very fortunate to be advised by Prof. Nakul Verma and Prof. Daniel Hsu.

I'm mainly interested in NLP—specifically, the principled understanding of self-supervised learning paradigms (text embeddings), self-attention models (transformers), and the geometric properties of representation learning (representation degeneration problem). I'm grateful to be supported by the Egleston Scholars Program and the Ezoe Memorial Recruit Foundation.

Email / CV / LinkedIn / Twitter / Github

Selected Publications

Do Models Explain Themselves? Counterfactual Simulatability of Natural Language Explanations
Yanda Chen, Ruiqi Zhong, Narutatsu Ri, Chen Zhao, He He, Jacob Steinhardt, Zhou Yu, Kathleen McKeown
arXiv 2023
[paper]

Contrastive Loss is All You Need to Recover Analogies as Parallel Lines
Narutatsu Ri, Fei-Tzin Lee, Nakul Verma
ACL 2023 RepL4NLP
[paper] [code] [slides]

Enhancing Few-shot Text-to-SQL Capabilities of Large Language Models: A Study on Prompt Design Strategies
Linyong Nan, Yilun Zhao, Weijin Zou, Narutatsu Ri, Jaesung Tae, Ellen Zhang, Arman Cohan, Dragomir Radev
arXiv 2023
[paper]

TA Experience

COMS 4771: Machine Learning (Summer 2022, Fall 2022 (Head TA), Spring 2023 (Head TA))

Miscellaneous

[05/15/2023] Survey report on \(k\)-means++
[04/13/2023] Quick writeup on Neural Tangent Kernels
[12/19/2022] Writeup on embeddability of word analogies on the \(n\)-sphere
[07/05/2022] Presentation on Reconstructing the Cascade of Language Processing in the Brain using the Internal Computations of a Transformer-based Language Model at IRCN (University of Tokyo)
[06/13/2022] Quick report on Deep Embedded Clustering
[06/04/2019] Paper on 文章中の潜在要素を考慮した対話システム
[04/30/2018] Some robots I've built in the past

Website design taken from Jon Barron.