Dongjae Jeon

Dongjae Jeon

Master's Student
Yonsei University

I am a first-year master's student in Computer Science at Yonsei University,
working in the Artificial Intelligence and Information Systems Lab with Professor Albert No.

I am broadly interested in generative models, currently focusing on three specific areas:

I have prior experience in Continual Learning applied to computer vision, specifically Object Detection (See Awards).
My primary goal is to deepen our understanding of how machines perceive and, ultimately, leverage that understanding to empower AI for societal good.

News

Jan 2026 Two papers on Diffusion Language Model (Rainbow Padding, A2D) is accepted to ICLR 2026. See you in Brazil 🇧🇷!
Nov 2025 A paper on Machine Unlearning (IDI) is accepted to AAAI 2026!
Oct 2025 Two preprints are now available: one on Diffusion Language Model training (Rainbow Padding) and Safety (A2D).
Sep 2025 A paper on Diffusion Language Model is accepted to NeurIPS 2025. Kudos to my co-authors! See you in San Diego 🇺🇸!
Aug 2025 Official code for ICML 2025 paper is uploaded! Checkout!
Aug 2025 I gave a talk at Yonsei MLSys student group on Diffusion models & their privacy issues.
Jun 2025 A paper on Quantization is accepted to ICML 2025 (TTODLer-FM workshop) as Oral!
May 2025 Two Papers (ODLRI and Regard-FT) are accepted to ACL 2025! See you in Austria 🇦🇹!

Publications

(*) denotes equal contribution

Preprints

Preserve then Quantize: Dominant-Subspace Guided Low-Rank Reconstruction
Preserve then Quantize: Dominant-Subspace Guided Low-Rank Reconstruction [P1]
Yoonjun Cho*, Dongjae Jeon*, Soeun Kim, Albert No
ICML (TTODLer-FM workshop), 2025 (Oral)
arXiv Slides

Peer-Reviewed

A2D: Any-Order, Any-Step Safety Alignment for Diffusion Language Models
A2D: Any-Order, Any-Step Safety Alignment for Diffusion Language Models [C6]
Wonje Jeung*, Sangyeon Yoon*, Yoonjun Cho, Dongjae Jeon, Sangwoo Shin, Hyesoo Hong, Albert No
ICLR, 2026
arXiv
Information-Theoretic Discrete Diffusion
Information-Theoretic Discrete Diffusion [C4]
Moongyu Jeon, Sangwoo Shin, Dongjae Jeon, Albert No
NeurIPS, 2025
arXiv
Assigning Distinct Roles to Quantized and Low-Rank Matrices Toward Optimal Weight Decomposition
Assigning Distinct Roles to Quantized and Low-Rank Matrices Toward Optimal Weight Decomposition [C3]
Yoonjun Cho, Soeun Kim, Dongjae Jeon, Kyelim Lee, Beomsoo Lee, Albert No
ACL Findings, 2025
arXiv
Understanding and Mitigating Memorization in Generative Models via Sharpness of Probability Landscapes
Understanding and Mitigating Memorization in Generative Models via Sharpness of Probability Landscapes [C2]
Dongjae Jeon*, Dueun Kim*, Albert No
ICML, 2025 (Spotlight)
Previously at AAAI (PPAI workshop), 2025 (Oral)
arXiv Code Slides
Large Language Models Still Exhibit Bias in Long Text
Large Language Models Still Exhibit Bias in Long Text [C1]
Wonje Jeung, Dongjae Jeon, Ashkan Yousefpour, Jonghyun Choi
ACL Findings, 2025
Previously at NeurIPS (SOLAR workshop), 2024
arXiv

Awards

Class-Incremental with Repetition (CIR) using Unlabelled Data
Class-Incremental with Repetition (CIR) using Unlabelled Data
🥈 2nd Place
CVPR CLVISION challenge, 2024
Continual Test-time Adaptation for Object Detection
Continual Test-time Adaptation for Object Detection
🥇 1st Place
ICCV VCL challenge, 2023
Slides Report

Talks

Privacy Issues in DGMs: How to detect & mitigate
Privacy Issues in DGMs: How to detect & mitigate
MLSys Student Group Seminar, 2025
Slides

Blog

Jan 2026

Coming Soon

...will be updated soon...

Read more →