Publications

* denotes equal contribution. Authorship in robust statistics is in alphabetical order.

Language Models

On-policy self-distillation via prompt optimization

Under review.

Looped diffusion language models

Sanghyun Lee, Chunsan Hong, Seungryong Kim, Jonghyun Lee, JHP, Dongmin Park

Under review.

ThinkSafe: self-generated safety alignment for reasoning models [slides]

Seanie Lee, Sangwoo Park, Yumin Choi, Gyeongman Kim, Minki Kang, Jihun Yun, Dongmin Park, JHP, Sung Ju Hwang

Under Review.

Effective test-time scaling of discrete diffusion through iterative refinement

Sanghyun Lee, Sunwoo Kim, Seungryong Kim, JHP, Dongmin Park

Under review. SPIGM workshop at ICML 2026.

Alignment as distribution learning: your preference model is explicitly a language model

Jihun Yun*, Juno Kim*, JHP, Jongha Jon Ryu, Junhyuck Kim, Jaewoong Cho, Kwang-Sung Jun

Under review. FoPT workshop at COLT 2025.

Lookahead unmasking elicits accurate decoding in diffusion language

Sanghyun Lee, Seungryong Kim, JHP, Dongmin Park

ICML 2026.

Not all bits are equal: scale-dependent memory optimization strategies for reasoning models

Junhyuck Kim, Ethan Ewer, Taehong Moon, JHP, Dimitris Papailiopoulos

ICLR 2026. ER workshop at NeurIPS 2025 (spotlight).

Can large language models develop strategic reasoning? Post-training insights from learning chess

Dongyoon Hwang, Hojoon Lee, Jaegul Choo, Dongmin Park, JHP

ScalR workshop at COLM 2025.

Lexico: extreme KV cache compression via sparse coding over universal dictionaries

Junhyuck Kim, JHP, Jaewoong Cho, Dimitris Papailiopoulos

ICML 2025. ICLR Workshop on Sparsity in LLMs (spotlight).

Task diversity shortens the ICL plateau

Jaeyeon Kim, Sehyun Kwon, Joo Young Choi, JHP, Jaewoong Cho, Jason D. Lee, Ernest K. Ryu

TMLR 2025.

Can Mamba learn how to learn? A comparative study on in-context learning tasks

JHP, Jaeseung Park, Zheyang Xiong, Nayoung Lee, Jaewoong Cho, Samet Oymak, Kangwook Lee, Dimitris Papailiopoulos

ICML 2024.

2023

Prompted LLMs as chatbot modules for long open-domain conversation

Gibbeum Lee*, Volker Hartmann*, JHP*, Dimitris Papailiopoulos, Kangwook Lee

Findings of ACL 2023.

2019

Literary event detection

Matthew Sims, JHP, David Bamman

ACL 2019.

Robust Statistics

2023

First order stochastic optimization with oblivious noise

Ilias Diakonikolas, Sushrut Karmalkar, JHP, Christos Tzamos

NeurIPS 2023.

GLM regression with oblivious corruptions

Ilias Diakonikolas, Sushrut Karmalkar, JHP, Christos Tzamos

COLT 2023.

ReLU regression with Massart noise

Ilias Diakonikolas, JHP, Christos Tzamos

NeurIPS 2021.

On robust mean estimation under coordinate-level corruption

Zifan Liu*, JHP*, Theodoros Rekatsinas, Christos Tzamos

ICML 2021.

Teaching

I have taught the following courses as a TA (*Head TA) in my undergraduate and graduate years.

UC Berkeley

  • Su17: CS70 (Discrete Math and Probability)

  • Fa18, Sp19: CS170 (Algorithms)

UW Madison

  • Fa19: CS240 (Discrete Math)

  • Sp20, Sp21*: CS577 (Algorithms)

  • Fa20: CS787 (Advanced Algorithms)