Justin Cho

조현동

PhD Candidate in Computer Science

hd [dot] justincho [at] gmail [dot] com / google scholar / resume / blog

👋 Hi there!

I am a PhD candidate at the University of Southern California's (USC) Information Sciences Institute (ISI), advised by Jonathan May.

My research is in natural language processing and machine learning, and most of my research have been centered around improving and understanding the role of language models in human-AI interactions. I've always been mesmerized by likeable and superintelligent digital assistants in science fiction such as Samantha in Her and TARS in Interstellar, and I'm excited to make them a reality.

Previously, I did my undergraduate studies at Hong Kong University of Science and Technology with a major in Computer Science, where I first worked on NLP with Pascale Fung. I have also interned at Amazon Alexa, working with Nicolaas Jedema, Alessandro Moschitti, and Pedro Szekely(2022), at Meta AI on Alborz Geramifard's Cognitive AI team with Chinnadhurai Sankar (2022) and with Ahmad Beirami (2021), Stitch Fix (2020), and at ISI's Natural Language Group with Jonathan May (2019).

📆 Office hours

I host virtual office hours! I've had the pleasure making many new connections through these office hours. I'm open to discussing anything from research to career advice that I can help with. Please schedule through Calendly. Email me if the time slots don't work for you.

🔍 Research

speechllm-thumbnail
Hyundong J. Cho, Nicolaas Jedema, Leonardo F. R. Ribeiro, Karishma Sharma, Pedro Szekely, Alessandro Moschitti, Ruben Janssen, Jonathan May. EMNLP 2024 [project page]
Current LLMs are fine-tuned with data exclusively with a text interface, which does not capture human preferences for speech, and thus generate text that is not suitable for text-to-speech systems. We collect 20K human preference data where annotators listen to the paired responses, instead of reading them through a text interface. We use this data for reinforcement learning with human feedback to adapt an instruction-tuned language model to generate speech-suitable text.
boteval-thumbnail
Hyundong J. Cho, Thamme Gowda, Yuyang Huang, Zixun Lu, Tianli Tong, Jonathan May. ACL 2024 SDT
We develop BotEval, an easily customizable, open-source, evaluation toolkit that focuses on enabling human-bot interactions as part of the evaluation process, as opposed to human evaluators evaluating a static input.
convmod-thumbnail
Hyundong J. Cho, Shuai Liu, Taiwei Shi, Darpan Jain, Basem Rizk, Yuyang Huang, Zixun Lu, Nuan Wen, Jonathan Gratch, Emilio Ferrera, Jonathan May. NAACL 2024 [project page]
Deleting comments and banning users are iron-fisted moderation tactics that can lead to a chilling effect on free speech. Instead, conversational moderation aims to guide users to more constructive behavior. We investigate whether language models can be effective conversational moderators and thus be effective tools for improving the health of online discourse.
dstegqa-thumbnail
Hyundong J. Cho, Andrea Madotto, Zhaojiang Lin, Khyathi Raghavi Chandu, Satwik Kottur, Jing Xu, Jonathan May, Chinnadhurai Sankar, EMNLP2023 [paper]
Estimating a user's goal in a dialogue can be done by asking natural language questions, and answering questions is a transferable skill that can be easily learned from examples. With this insight, we restructure dialogue state tracking (DST) to eliminate service-specific structured text and unify data from all services by decomposing each DST sample to a bundle of fine-grained example-guided question answering tasks. With a retriever trained to find examples that introduce similar updates to dialogue states, we find that our method can significantly boost continual learning performance, even for a model with just 60M parameters.
normvio-thumbnail
Jihyung Moon, Dong-Ho Lee, Hyundong J. Cho, Woojeong Jin, Chan Young Park, Minwoo Kim, Jonathan May, Jay Pujara, Sungjoon Park, EMNLP2023 [paper]
Toxic behavior in live-stream chat is a growing concern as live-streaming platforms such as Twitch and YouTube live are becoming increasingly popular. Previous detection methods are not effective for live-stream chat as each comment is only visible for a limited time and lacks a thread structure. To bridge this gap, we define norm violation categories in live-stream chats and annotate 4,583 moderated comments from Twitch and train live-stream chat-specific detection models.
recap-thumbnail
Shuai Liu, Hyundong J. Cho, Marjorie Freedman, Xuezhe Ma, Jonathan May, ACL2023 [paper]
Endowing chatbots with a consistent persona is essential to an engaging conversation, yet it remains an unresolved challenge. In this work, we propose a new retrieval-enhanced approach for personalized response generation.
checkdst-thumbnail
Hyundong J. Cho, Chinnadhurai Sankar, Christopher Lin, Kaushik Ram Sadagopan, Shahin Shayandeh, Asli Celikyilmaz, Jonathan May, Ahmad Beirami, EMNLP2022 Findings [paper] [project page]
Humans are robust to understanding dialogue states in the presence of noise and ambiguity, but dialogue state tracking (DST) models are not. This analysis of DST robustness has been sparse and uncoordinated in previous work. Our standardized and comprehensive DST diagnoses toolkit, CheckDST, is a collection of robustness tests and failure mode analytics. With CheckDST, we discover that different classes of DST models have clear strengths and weaknesses, where generation models are more promising for handling language variety while classification models are more robust to unseen entities.
reflect-thumbnail
Pei Zhou, Hyundong J. Cho, Pegah Jandaghi, Dong-Ho Lee, Bill Yuchen Lin, Jay Pujara, Xiang Ren. EMNLP2022 [paper] [project page]
We introduce Reflect, a dataset that annotates dialogues with explicit CG and solicits 9k diverse human-generated responses each following one common ground. Using Reflect, we showcase the limitations of current dialogue data and RG models: less than half of the responses in current data are rated as high quality and models trained using this data have even lower quality, while most Reflect responses are judged high quality. We also analyze whether CG can help models produce better-quality responses by using Reflect CG to guide RG models.
probing-thumbnail
Pei Zhou, Pegah Jandaghi, Hyundong J. Cho, Bill Yuchen Lin, Jay Pujara, Xiang Ren. EMNLP2021 Findings [paper]
We collect 6k annotated explanations justifying responses from four dialogue datasets and ask humans to verify them and propose two probing settings to evaluate response generation models' commonsense reasoning capabilities. Probing results show that response generation models fail to capture the logical relations between commonsense explanations and responses and fine-tuning on in-domain data and increasing model sizes do not lead to understanding of commonsense reasoning.
spolin-thumbnail
Open-domain dialogue systems overlook an important phenomena that makes conversations engaging: the initiation of the next relevant contribution, which is the most proactive method of "grounding". We collect "Yes, and" type dialogue pairs that naturally embed such initations from an improv podcast and existing dialogue corpora to create the Spontaneanation Pairs Of Learnable ImprovisatioN (SPOLIN) dataset. Human evaluation shows that models fine-tuned with SPOLIN generate more engaging results.

🔍 Others

viola-thumbnail
Hyundong J. Cho, Basel Shbita, Kartik Shenoy, Shuai Liu, Nikhil Patel, Hitesh Pindikanti, Jennifer Lee, Jonathan May. Alexa Prize Socialbot Grand Challenge 4 Proceedings, 2021
We present Viola, an open-domain dialogue system based on a simple generate-and-rank approach. Viola fetches a batch of response candidates from various neural dialogue models and template-based generators and chooses the final response with a poly-encoder ranker fine-tuned with annotated Alexa conversation data.

📢 News

  • 2024/9: Our work "Speechworthy Instruction-tuned Language Models" has been accepted to EMNLP 2024! See you in Miami!
  • 2024/6: Our work "BotEval: Facilitating Interactive Human Evaluation" has been accepted to ACL 2024 SDT! I've also started my internship with Amazon as an applied scientist intern.
  • 2024/3: Our work "Can Language Model Moderators Improve the Health of Online Discourse?" has been accepted to NAACL 2024! See you in Mexico City!
  • 2024/2: I'm giving a talk at Tonal on "Large Language Models: The Recipes and Applications of ChatGPT's Engine" to give an overview of recent advances in NLP and my work on Speechworthy Instruction-tuned Language Models.
  • 2023/11: I'll be attending EMNLP 2023 in person to present Continual Dialogue State Tracking via Example-Guided Question Answering and Analyzing Norm Violations in Live-Stream Chat.
  • 2023/7: I'm an organizer for ICML 2023's What's left to TEACH chatbots? workshop. See you in Hawaii!
  • 2023/6: I've started my internship with Amazon Alexa AI to work on adapting language models for voice-based interactions using reinforcement learning with human feedback.
  • 2022/11: I'll be attending EMNLP 2022 in person to present Know Thy Strengths: Comprehensive Dialogue State Tracking Diagnostics and Reflect, Not Reflex: Inference-Based Common Ground Improves Dialogue Response Quality.
  • 2022/6: I've started my summer internship at Meta AI to work on continual learning.
  • 2021/8: I started my internship with Meta AI's Conversational AI team to work on the robustness of task-oriented dialogue models.
  • 2021/8: "Probing Causal Common Sense in Dialogue Response Generation", work with Pei Zhou, has been accepted to EMNLP2021 Findings.
  • 2021/4: Viola makes it to the semi-finals of the Alexa Prize Socialbot Grand Challenge 4!
  • 2020/11: Viola is one of the teams accepted to compete in the Alexa Prize Socialbot Grand Challenge 4! I will be leading the team with Jonathan May as our faculty advisor.
  • 2020/9: Stitch Fix posted a blog post about my internship project using NLP to process client feedback for its products. Check it out!
  • 2020/9: USC Viterbi Magazine covered Jonathan May's and my SPOLIN work that was published at ACL2020, with a fun video demonstration featuring Mike Henry (Family Guy, The Orville). Check it out!
  • 2020/8: I am starting my first semester as a PhD student at USC.
  • 2020/6: I will be working at Stitch Fix as a data science intern in the merch product development team.
  • 2020/4: My paper with Jonathan May has been accepted to ACL2020! I will be presenting virtually at the conference.

➕ Miscellaneous

  • My pronouns are he, him, his.
  • My dad being a diplomat, I have been blessed with the experience of living in many different countries: Moscow, Russia; Oslo, Norway; Vienna, Austria; Abu Dhabi, UAE; Hong Kong; Tehran, Iran; Seoul, South Korea; Los Angeles, California. Unfortunately, I haven't lived long enough in most of these countries to master a new language other than Korean and English. My favorite country to live in was Austria, where I spent 5-7th grade. I loved being in central Europe and being able to easily travel to nearby countries.
  • I'm working on a language learning app called AutoLang with a few colleagues from college. I was frustrated that I couldn't say what I wanted to say to my Chinese peers even after mastering all the Chinese lessons on Duolingo. I realized I need a more personalized, bottom-up approach and decided to make one myself with a couple of friends. With AutoLang, you start learning things relevant to you from day 1. It's designed to help you build confidence to converse by talking to discover things you didn't know how to say or understand and automatically build your own curriculum with flashcards you create in the process! It's not beginner friendly yet and more suited to those who are already at an intermediate level, but we're working on making it more usable for beginners too.
  • I'm a bit of a productivity geek. Ask me about knowledge management and productivity tips! I use Logseq for knowledge management and use tips from How to Take Smart Notes, which recommends the Zettelkasten method. I use Brick for making my phone a dumb phone and Screentime for blocking distractions on Chrome.
  • I love playing football⚽ and I am a huge fan of FC Barcelona. Visca el Barça!
  • I am a citizen of South Korea and the US. I am fluent in both English and Korean. I am teaching myself Mandarin with AutoLang and HSK level tests. 我希望能在不远的将来用流利的中文和你交流。