KIWAN MAENG

Publications

(Systems for) Privacy-preserving Machine Learning

• Correlating Cross-Iteration Noise for DP-SGD using Model Curvature

  NeurIPS 2026 [PDF]
  Xin Gu, Yingtai Xiao, Guanline He, Jiamu Bai, Daniel Kifer, and Kiwan Maeng

• Spectrum-Aware Bounds on Invertibility for Privacy-Enhancing Instance Encoding

  Arxiv [PDF]
  Seokjin Hwang*, Yuting (Ray) Li*, and Kiwan Maeng

• Composition for Pufferfish Privacy

  VLDB 2026 [PDF]
  Jiamu Bai, Guanlin He, Xin Gu, Daniel Kifer, and Kiwan Maeng

• Cocoon: A System Architecture for Differentially Private Training with Correlated Noises

  OSDI 2026 (also presented at 2026 OCP Global Summit) [PDF]
  Donghwan Kim, Xin Gu, Jinho Baek, Timothy Lo, Younghoon Min, Kwangsik Shin, Jongryool Kim, Jongse Park, and Kiwan Maeng

• Beyond Latency: A System-Level Characterization of MPC and FHE for PPML

  ISPASS 2026 [PDF]
  Pengzhi Huang, Kiwan Maeng, and G. Edward Suh

• CrypTorch: PyTorch-based Auto-tuning Compiler for Machine Learning with Multi-party Computation

  Arxiv [PDF]
  Jinyu Liu, Gang Tan, and Kiwan Maeng

• In-depth Characterization of Machine Learning on an Optimized Multi-party Computing Library

  CAL 2025 [PDF]
  Jinyu Liu and Kiwan Maeng

• Characterizing the System Overhead of Discrete Noise Generation for Differential Privacy

  CAL 2025 [PDF]
  Seokhyeon Kong, Donghwan Kim, Euiseong Seo, and Kiwan Maeng

• Practical Federated Recommendation Model Learning Using ORAM with Controlled Privacy

  ASPLOS 2025 [PDF]
  Jinyu Liu, Wenjie Xiong, G. Edward Suh, and Kiwan Maeng

• Information Flow Control in Machine Learning through Modular Model Architecture

  Usenix Security 2024 [PDF]
  Trishita Tiwari, Suchin Gururangan, Chuan Guo, Weizhe Hua, Sanjay Kariyappa, Udit Gupta, Wenjie Xiong, Kiwan Maeng, Hsien-Hsin S. Lee, and G. Edward Suh

• Accelerating ReLU for MPC-Based Private Inference with a Communication-Efficient Sign Estimation

  MLSys 2024 [PDF]
  Kiwan Maeng and G. Edward Suh

• Compiler-based Memory Encryption for Machine Learning on Commodity Low-power Devices

  CC 2024 [PDF]
  Kiwan Maeng and Brandon Lucia

• LazyDP: Co-Designing Algorithm-Software for Scalable Training of Differentially Private Recommendation Models

  ASPLOS 2024 [PDF]
  Juntaek Lim, Youngeun Kwon, Ranggi Hwang, Kiwan Maeng, G. Edward Suh, and Minsoo Rhu

• Bounding the Invertibility of Privacy-preserving Instance Encoding using Fisher Information

  NeurIPS 2023 [PDF]
  Kiwan Maeng*, Chuan Guo*, Sanjay Kariyappa, and G. Edward Suh

• GPU-based Private Information Retrieval for On-Device Machine Learning Inference

  ASPLOS 2024 [PDF]
  Maximilian Lam, Jeff Johnson, Wenjie Xiong, Kiwan Maeng, Udit Gupta, Minsoo Rhu, Hsien-Hsin S. Lee, Vijay Janapa Reddi, Gu-Yeon Wei, David Brooks, and G. Edward Suh

• Cocktail Party Attack: Breaking Aggregation-Based Privacy in Federated Learning using Independent Component Analysis

  ICML 2023 [PDF]
  Sanjay Kariyappa, Chuan Guo, Kiwan Maeng, Wenjie Xiong, G. Edward Suh, Moinuddin K Qureshi, and Hsien-Hsin S. Lee

• Measuring and Controlling Split Layer Privacy Leakage Using Fisher Information

  FL-NeurIPS 2022 Workshop [PDF]
  Kiwan Maeng, Chuan Guo, Sanjay Kariyappa, and G. Edward Suh

• Towards Fair Federated Recommendation Learning: Characterizing the Inter-Dependence of System and Data Heterogeneity

  RecSys 2022 (Best Paper Finalist) [PDF]
  Kiwan Maeng, Haiyu Lu, Luca Melis, John Nguyen, Mike Rabbat, and Carole-Jean Wu

• Sustainable AI: Environmental Implications, Challenges and Opportunities

  MLSys 2022 [PDF]
  Carole-Jean Wu, Ramya Raghavendra, Udit Gupta, Bilge Acun, Newsha Ardalani, Kiwan Maeng, Gloria Chang, Fiona Aga Behram, James Huang, Charles Bai, Michael Gschwind, Anurag Gupta, Myle Ott, Anastasia Melnikov, Salvatore Candido, David Brooks, Geeta Chauhan, Benjamin Lee, Hsien-Hsin S. Lee, Bugra Akyildiz, Maximilian Balandat, Joe Spisak, Ravi Jain, Mike Rabbat, and Kim Hazelwood

• Green Federated Learning

  FL-ICML 2023 Workshop [PDF]
  Ashkan Yousefpour, Shen Guo, Ashish Shenoy, Sayan Ghosh, Pierre Stock, Kiwan Maeng, Schalk-Willem Krüger, Michael Rabbat, Carole-Jean Wu, and Ilya Mironov

• FEL: High Capacity Learning for Recommendation and Ranking via Federated Ensemble Learning

  IEEE Data Eng. Bull 2023 [PDF]
  Meisam Hejazinia, Dzmitry Huba, Ilias Leontiadis, Kiwan Maeng, Mani Malek, Luca Melis, Ilya Mironov, Milad Nasr, Kaikai Wang, and Carole-Jean Wu

Systems for ML

• Characterizing How Complex Agentic AI Systems Handle General Tasks: A Trace-Based Simulation Study

  IISWC 2026 (Best Paper Finalist) [PDF]
  Donghwan Kim, Prakhar Singh, Younghoon Min, Jongryool Kim, Jongse Park, and Kiwan Maeng

• Hermes: Algorithm-System Co-design for Efficient Retrieval Augmented Generation At-Scale

  ISCA 2025 [PDF]
  Michael Shen, Muhammad Umar, Kiwan Maeng, G. Edward Suh, and Udit Gupta

• Towards Understanding Systems Trade-offs in Retrieval-Augmented Generation Model Inference

  Arxiv [PDF]
  Michael Shen, Muhammad Umar, Kiwan Maeng, G. Edward Suh, and Udit Gupta

• Pirate: No Compromise Low-Bandwidth VR Streaming for Edge Devices

  ASPLOS 2025 [PDF]
  Yingtian Zhang, Yan Kang, Ziyu Ying, Wanhang Lu, Sijie Lan, Huijuan Xu, Kiwan Maeng, Anand Sivasubramaniam, Mahmut Kandemir, and Chita Das

• SVDE: Serverless Framework for Low-Latency Video Analytic Queries With Hardware Disaggregation

  IEEE Transactions on Computers 2025 (Best Paper Award) [PDF]
  Pingyi Huo, Theodore Michailidis, Yi Zheng, Prapti Panigrahi, Kiwan Maengu, Jishen Zhao, and Vijaykrishnan Narayanan

• Optimizing CPU Performance for Recommendation Systems At-Scale

  ISCA 2023 [PDF]
  Rishabh Jain, Scott Cheng, Vishwas Kalagi, Vrushabh Sanghavi, Samvit Kaul, Meena Arunachalam, Kiwan Maeng, Adwait Jog, Anand Sivasubramaniam, Mahmut Taylan Kandemir, and Chita Das

• Carbon Explorer: A Holistic Approach for Designing Carbon Aware Datacenters

  ASPLOS 2023 [PDF]
  Bilge Acun, Benjamin Lee, Fiodar Kazhamiaka, Kiwan Maeng, Manoj Chakkaravarthy, Udit Gupta, David Brooks, and Carole-Jean Wu

• Carbon Dependencies in Datacenter Design and Management

  HotCarbon 2022 [PDF]
  Bilge Acun, Benjamin Lee, Fiodar Kazhamiaka, Aditya Sundarrajan, Manoj Chakkaravarthy, Kiwan Maeng, David Brooks, and Carole-Jean Wu

• CPR: Understanding and Improving Failure Tolerant Training for Deep Learning Recommendation with Partial Recovery

  MLSys 2021 [PDF]
  Kiwan Maeng, Shivam Bharuka, Isabel Gao, Mark C. Jeffrey, Vikram Saraph, Bor-Yiing Su, Caroline Trippel, Jiyan Yang, Mike Rabbat, Brandon Lucia, and Carole-Jean Wu

Batteryless Energy-harvesting Devices

• An Architectural Charge Management Interface for Energy-Harvesting Systems

  MICRO 2022 [PDF]
  Emily Ruppel, Milijana Surbatovich, Harsh Desai, Kiwan Maeng, and Brandon Lucia

• Adaptive Low-overhead Scheduling for Periodic and Reactive Intermittent Execution

  PLDI 2020 [PDF]
  Kiwan Maeng and Brandon Lucia

• Enhancing Stratospheric Weather Analysis and Forecasts by Deploying Sensors from a Weather Balloon

  NeurIPS 2019 Workshop (Spotlight Talk) [PDF]
  Kiwan Maeng, Iskender Kushan, Brandon Lucia, and Ashish Kapoor

• Supporting Peripherals in Intermittent Systems with Just-In-Time Checkpoints

  PLDI 2019 [PDF]
  Kiwan Maeng and Brandon Lucia

• Adaptive Dynamic Checkpointing for Safe Efficient Intermittent Computing

  OSDI 2018 [PDF]
  Kiwan Maeng and Brandon Lucia

• Alpaca: Intermittent Execution Without Checkpoints

  OOPSLA 2017 [PDF] [Ext. ver.]
  Kiwan Maeng, Alexei Colin, and Brandon Lucia

• Intermittent Computing: Challenges and Opportunities

  SNAPL 2017 [PDF]
  Brandon Lucia, Vignesh Balaji, Alexei Colin, Kiwan Maeng, and Emily Ruppel

Services

• Organizing: MICRO 2025 Student Travel Grant Co-Chair, ASPLOS 2025 Artifact Evaluation Committee Co-Chair
• Session Chair: ASPLOS 2024 (3D: ML Quantization and Memory Optimizations), ASPLOS 2026 (1B: LLM Serving: Latency & Scheduling)
• Conference (External) Program Committee: ASPLOS (2023-2027), PACT (2026), MLSys (2022-2025), HPCA (2024-2025), ISCA (2024), MICRO (2023, 2025-2026), EuroSys (2022), IISWC (2022)
• Conference Reviewer: NeurIPS (2025; Best reviewer)
• Journal Reviewer: Transactions on Storage (2025-2026), IEEE Computer Architecture Letters (2025-2026), IEEE Transactions of Computers (2025), IEEE Micro (2022), ACM TACO (2022)
• Workshop Program Committee: ASPLOS YArch (2024), PACT SRC (2022), SenSys USN (2022)
• NSF Panel: CCF-SHF (2024)

Awards & Honors

• NSF SHF: Practical Privacy-preserving AI Offloading with Hybridizing Multi-party Computing and Strongly-secure Hardware (2025 -- 2027; $600K), Single PI
• NSF SaTC: Utilizing Untrusted Machine Learning Services Safely via Privacy-Preserving User Data Embedding (2024 -- 2026; $600K), PI (co-PI prof. Kifer)
• Charles K. Etner Early Career Professorship (2022 -- 2025)
• KFAS Scholarship (2016 -- 2021)
• Summa Cum Laude (B.S., Seoul National University)
• National Scholarship for Science and Engineering (2010 -- 2016)

Industry Experiences

• Postdoc Researcher. Facebook AI Research (FAIR; 2021 -- 2022)
• Research Intern. Facebook AI Research (FAIR; 2020.5 -- 2020.12)
• Research Intern. Microsoft Research (2019.5 -- 2019.8)

Miscellaneous

I used to be an amateur webtoon--(web + cartoon)--ist. I used to draw cartoons about scientists and engineers on Facebook, and they were popular enough to become a real book! Do not bother clicking the link if you cannot read Korean :)
I haven't drawn any new episodes in years, but still I have things to boast:

• I was awarded as 10 Authors of the Year 2017 by Chosun Ilbo (largest Korean newspaper).
• My book was awarded as 10 Books of the Year 2017 by Munhwa Ilbo (newspaper), Books of the Year 2017 by Aladin (bookstore), Recommendations for Students By Librarians 2017 by Jeju Ilbo (newspaper), ranked 8th in Books of the Year Selected by Publishers SisaIN (magazine), and was nominated as Book of The Year by Donga Ilbo (newspaper).
• My book was selected as Sejong-Books 2017 by Publication Industry Promotion Agency of Korea. No idea what it is but doesn't it sound fancy?
• I donated all my profit from my first print in the hope to ease starvation in the République du Niger. Donation was made through Save The Children Federation, Inc.
• My book was selected and now can be found in libraries of the Korean military (Republic of Korea Armed Forces), a.k.a., Jin-joong-moon-go.
• I used to have 58 thousand followers on Facebook when Facebook was a thing.