Yuntian Deng

About

I am an assistant professor at the University of Waterloo. My research interests are Natural Language Processing and Machine Learning. I received my PhD from Harvard University, where I was advised by Prof. Alexander Rush and Prof. Stuart Shieber. I did a postdoc under the supervision of Prof. Yejin Choi.

Work

University of Waterloo
|

Assistant Professor

Canada

University of Waterloo
|

Assistant Professor

Canada

→

Allen Institute for Artificial Intelligence
|

Postdoc

US

→

Nvidia (United States)
|

Intern

US

→

Facebook (United States)
|

Intern

US

→

Bloomberg (United States)
|

Intern

US

→

Education

Carnegie Mellon University
United States of America

→

Master of Science in Language Technologies

Tsinghua University
China

→

Bachelor of Engineering

Harvard University
United States of America

→

PhD in Computer Science

Publications

Discrete Transformer

Summary

journal-article

Magpie: Alignment Data Synthesis from Scratch by Prompting Aligned LLMs with Nothing

Published by

arXiv preprint arXiv:2406.08464

Summary

journal-article

MixEval: Deriving Wisdom of the Crowd from LLM Benchmark Mixtures

Published by

arXiv preprint arXiv:2406.06565

Summary

journal-article

WILDBENCH: Benchmarking LLMs with Challenging Tasks from Real Users in the Wild

Published by

arXiv preprint arXiv:2406.04770

Summary

journal-article

WildChat: 1M ChatGPT Interaction Logs in the Wild

Published by

ICLR

Summary

conference-paper

WildHallucinations: Evaluating Long-form Factuality in LLMs with Real-World Entity Queries

Published by

arXiv preprint arXiv:2407.17468

Summary

journal-article

WildVis: Open Source Visualizer for Million-Scale Chat Logs in the Wild

Published by

arXiv preprint arXiv:2409.03753

Summary

journal-article

From explicit cot to implicit cot: Learning to internalize cot step by step

Published by

arXiv preprint arXiv:2405.14838

Summary

journal-article

GenSLMs: Genome-scale language models reveal SARS-CoV-2 evolutionary dynamics

Published by

The International Journal of High Performance Computing Applications

Summary

journal-article

Implicit chain of thought reasoning via knowledge distillation

Published by

arXiv preprint arXiv:2311.01460

Summary

journal-article

Instruction in the wild: A user-based instruction dataset

Published by

GitHub repository

Summary

journal-article

Structure Modeling for Language Models

Summary

dissertation-thesis

Tree Prompting: Efficient Task Adaptation without Fine-Tuning

Published by

EMNLP

Summary

conference-paper

DeepSpeed4Science Initiative: Enabling Large-Scale Scientific Discovery through Sophisticated AI System Technologies

Published by

arXiv preprint arXiv:2310.04610

Summary

journal-article

Model Criticism for Long-Form Text Generation

Published by

EMNLP

Summary

conference-paper

Semi-Parametric Inducing Point Networks and Neural Processes

Published by

ICLR

Summary

conference-paper

Markup-to-Image Diffusion Models with Scheduled Sampling

Published by

ICLR

Summary

conference-paper

Rationales for Sequential Predictions

Published by

EMNLP

Summary

conference-paper

Sequence-to-Lattice Models for Fast Translation

Published by

EMNLP Findings

Summary

conference-paper

Low-Rank Constraints for Fast Inference in Structured Models

Published by

NeurIPS

Summary

conference-paper

Residual Energy-Based Models for Text Generation

Published by

ICLR

Summary

conference-paper

Cascaded Text Generation with Markov Transformers

Published by

NeurIPS

Summary

conference-paper

Challenges in end-to-end neural scientific table recognition

Published by

ICDAR

Summary

conference-paper

Neural Linguistic Steganography

Published by

EMNLP

Summary

conference-paper

Real or Fake? Learning to Discriminate Machine from Human Generated Text

Published by

arXiv preprint arXiv:1906.03351

Summary

journal-article

AdaptivFloat: A Floating-point based Data Type for Resilient Deep Learning Inference

Published by

DAC

Summary

conference-paper

Latent Alignment and Variational Attention

Published by

NeurIPS

Summary

conference-paper

Visual Attention Model for Cross-sectional Stock Return Prediction and End-to-End Multimodal Market Representation Learning

Published by

FLAIRS

Summary

conference-paper

Bottom-Up Abstractive Summarization

Published by

EMNLP

Summary

conference-paper

Learning Latent Space Models with Angular Constraints

Published by

ICML

Summary

conference-paper

OpenNMT: Open-Source Toolkit for Neural Machine Translation

Published by

ACL Demo

Summary

conference-paper

Image-to-Markup Generation with Coarse-to-Fine Attention

Published by

ICML

Summary

conference-paper

Neural Machine Translation with Recurrent Attention Modeling

Published by

EACL

Summary

conference-paper

Learning Concept Taxonomies from Multi-modal Data

Published by

ACL

Summary

conference-paper

Dropout with Expectation-linear Regularization

Published by

ICLR

Summary

conference-paper

On the generalization error bounds of neural networks under diversity-inducing mutual angular regularization

Published by

arXiv preprint arXiv:1511.07110

Summary

journal-article

Entity Hierarchy Embedding

Published by

ACL

Summary

conference-paper

Diversifying Restricted Boltzmann Machine for Document Modeling

Published by

KDD

Summary

conference-paper

Latent variable modeling with diversity-inducing mutual angular regularization

Published by

arXiv preprint arXiv:1512.07336

Summary

journal-article

Yuntian Deng