Publications

Publications are grouped by category in reverse chronological order. Preprints are ordered by the date first posted; revision dates are noted separately.

^* Equal contribution.

Journal articles · Conference papers · Preprints · Ph.D. thesis

Journal Articles

• Benign overfitting beyond prediction: The ordinary least squares interpolator
  Dennis Shen^*, Dogyoon Song^*, Peng Ding and Jasjeet S. Sekhon
  Biometrika, 2026. arXiv

• Local minima structures in Gaussian mixture models
  Yudong Chen^*, Dogyoon Song^*, Xumei Xi and Yuqian Zhang
  IEEE Transactions on Information Theory, 2024

• On approximations of the PSD cone by a polynomial number of smaller-sized PSD cones
  Dogyoon Song and Pablo A. Parrilo
  Mathematical Programming, 2023

• Kronecker-structured covariance models for multiway data
  Yu Wang, Zeyu Sun, Dogyoon Song and Alfred O. Hero
  Statistics Surveys, 2022

• On robustness of principal component regression
  Anish Agarwal^*, Devavrat Shah, Dennis Shen^* and Dogyoon Song^*
  Journal of the American Statistical Association (JASA), 2021; Earlier conference version: NeurIPS 2019 (oral)

• Nearest neighbors for matrix estimation interpreted as blind regression for latent variable model
  Yihua Li, Devavrat Shah, Dogyoon Song^* and Christina Lee Yu^*
  IEEE Transactions on Information Theory, 66(3):1760–1784, 2020; Earlier conference version: NeurIPS 2016

Conference Proceedings

• Design-based finite-sample analysis for regression adjustment
  Dogyoon Song
  AISTATS 2026. arXiv

• Scaling laws of signSGD in linear regression: When does it outperform SGD?
  Jihwan Kim, Dogyoon Song and Chulhee Yun
  ICLR 2026. arXiv

• High-dimensional sequential change detection
  Robert Malinas, Dogyoon Song, Benjamin D. Robinson and Alfred O. Hero
  IEEE ISIT 2025. arXiv

• Sequential community detection in high-dimensional temporal graphs
  Robert Malinas, Dogyoon Song and Alfred O. Hero
  Asilomar Conference on Signals, Systems, and Computers, 2024

• Improving training efficiency of diffusion models via multi-stage framework and tailored multi-decoder architecture
  Huijie Zhang^*, Yifu Lu^*, Ismail Alkhouri, Saiprasad Ravishankar, Dogyoon Song and Qing Qu
  CVPR 2024

• Efficient low-dimensional compression of overparameterized models
  Soo Min Kwon^*, Zekai Zhang^*, Dogyoon Song, Laura Balzano and Qing Qu
  AISTATS 2024

• Errors-in-variables Fréchet regression with low-rank covariate approximation
  Dogyoon Song^* and Kyunghee Han^*
  NeurIPS 2023

• Minimum-risk recalibration of classifiers
  Zeyu Sun^*, Dogyoon Song^* and Alfred O. Hero
  NeurIPS 2023 (spotlight)

• Community detection in high-dimensional graph ensembles
  Robert Malinas, Dogyoon Song and Alfred O. Hero
  Asilomar Conference on Signals, Systems, and Computers, 2023. arXiv

• Robustness-preserving lifelong learning via dataset condensation
  Jinghan Jia, Yihua Zhang, Dogyoon Song, Sijia Liu and Alfred O. Hero
  ICASSP 2023

• Sample-efficient reinforcement learning via low-rank matrix estimation
  Devavrat Shah, Dogyoon Song^*, Zhi Xu^* and Yuzhe Yang
  NeurIPS 2020

• On robustness of principal component regression
  Anish Agarwal^*, Devavrat Shah, Dennis Shen^* and Dogyoon Song^*
  NeurIPS 2019 (oral). Journal version: JASA, 2021

• Blind regression: Nonparametric regression for latent variable models via collaborative filtering
  Christina Lee^*, Yihua Li, Devavrat Shah and Dogyoon Song^*
  NeurIPS 2016. Journal version: IEEE Transactions on Information Theory, 2020

• Verifying start-up failures in coupled ring oscillators in presence of variability using predictive global optimization
  Taehwan Kim, Dogyoon Song, Sangho Youn, Jaejin Park, Hojin Park and Jaeha Kim
  IEEE/ACM ICCAD 2013

• Discretization and discrimination methods for design, verification, and testing of analog/mixed-signal circuits
  Jaeha Kim, Jiho Lee, Dogyoon Song, Taehwan Kim, Kyunghoon Kim, Seobin Jung and Sangho Youn
  IEEE CICC 2013

• A low-power high-radix switch fabric based on low-swing signaling and partially-activated input lines
  Dogyoon Song and Jaeha Kim
  VLSI-DAT 2013

Preprints

• Neumann-series corrections for regression adjustment in randomized experiments
  Dogyoon Song
  arXiv:2511.08539, November 2025

• On separability of covariance in multiway data analysis
  Dogyoon Song and Alfred O. Hero
  arXiv:2302.02415, 2023; revised March 2026

• Learning RUMs: Reducing mixture to single component via PCA
  Devavrat Shah and Dogyoon Song
  arXiv:1812.11917, 2018; revised March 2020

• Deconvolution with unknown error distribution interpreted as blind isotonic regression
  Devavrat Shah and Dogyoon Song
  arXiv:1803.03469, 2018; revised April 2020; presented at the Allerton Conference 2018

Ph.D. Thesis

• Addressing missing data and scalable optimization for data-driven decision making
  Ph.D. Thesis, Massachusetts Institute of Technology, June 2021
  Advised by Pablo A. Parrilo and Devavrat Shah