prof_pic.jpg

I am an Assistant Professor at KAIST Graduate School of AI, where I direct the Structured and Probabilistic Machine Learning (SPML) Lab.

My research focuses on developing machine learning algorithms for molecules, with applications to drug discovery and material design. I enjoy bringing a machine learning perspective—especially a probabilistic one—to scientific problems, and diving deep into the underlying physics, chemistry, and biology.

Members

Yunhui Jang, Hyosoon Jang, Hyomin Kim, Seonghyun Park, Seongsu Kim, Kiyoung Seong, Minkyu Kim, Dongyeop Woo, Nayoung Kim, Minsu Kim (Postdoc), Taewon Kim, Hyunjin Seo, Yinhua Piao (Postdoc), Yoonho Kim, Honghui Kim (Postdoc)

Alumni: Haeji Ko, Juwon Hwang

Publications

  1. Boltz is a Strong Baseline for Atom-level Representation Learning
    Hyosoon Jang, Hyunjin Seo, Yunhui Jang, Seonghyun Park, and Sungsoo Ahn
    [arxiv]
  2. Riemannian MeanFlow
    Dongyeop Woo, Marta Skreta, Seonghyun Park, Kirill Neklyudov, and Sungsoo Ahn
    [arxiv]
  3. Progressive Multi-Agent Reasoning for Biological Perturbation Prediction
    Hyomin Kim, Sang-Yeon Hwang, Jaechang Lim, Yinhua Piao, Yunhak Oh, Woo Youn Kim, Chanyoung Park, Sungsoo Ahn, and Junhyeok Jeon
    [arxiv]
  4. AtomMOF: All-Atom Flow Matching for MOF-Adsorbate Structure Prediction
    Nayoung Kim, Honghui Kim, Sihyun Yu, Minkyu Kim, Seongsu Kim, and Sungsoo Ahn
    [arxiv]
  5. CatFlow: Co-generation of Slab-Adsorbate Systems via Flow Matching
    Minkyu Kim, Nayoung Kim, Honghui Kim, and Sungsoo Ahn
    [arxiv]
  6. INDIBATOR: Diverse and Fact-Grounded Individuality for Multi-Agent Debate in Molecular Discovery
    Yunhui Jang, Seonghyun Park, Jaehyung Kim, and Sungsoo Ahn
    [arxiv]
  7. Latent Veracity Inference for Identifying Errors in Stepwise Reasoning (ICLR 2026)
    Minsu Kim, Jean-Pierre R. Falet, Oliver Ethan Richardson, Xiaoyin Chen, Moksh Jain, Sungjin Ahn, Sungsoo Ahn, and Yoshua Bengio
    [arxiv]
  8. Learning Flexible Forward Trajectories for Masked Molecular Diffusion (ICLR 2026)
    Hyunjin Seo*, Taewon Kim*, Sihyun Yu, and Sungsoo Ahn
    [arxiv / project]
  9. Learning Collective Variables from BioEmu with Time-Lagged Generation (ICLR 2026)
    Seonghyun Park, Kiyoung Seong, Soojung Yang, Rafael Gomez-Bombarelli, and Sungsoo Ahn
    [arxiv]
  10. DNACHUNKER: Learnable Tokenization for DNA Language Models
    Taewon Kim, Jihwan Shin, Hyomin Kim, Youngmok Jung, Jonhoon Lee, Won-Chul Lee, Insu Han†, and Sungsoo Ahn†
    [arxiv]
  11. Energy-based Generator Matching: A Neural Sampler for General State Space (NeurIPS 2025)
    Dongyeop Woo, Minsu Kim, Minkyu Kim, Kiyoung Seong, and Sungsoo Ahn
    [arxiv / code]
  12. On Scalable and Efficient Training of Diffusion Samplers (NeurIPS 2025)
    Minkyu Kim, Kiyoung Seong, Dongyeop Woo, Sungsoo Ahn, and Minsu Kim
    [arxiv / code]
  13. Flexible MOF Generation with Torsion-Aware Flow Matching (NeurIPS 2025)
    Nayoung Kim, Seongsu Kim, and Sungsoo Ahn
    [arxiv / code]
  14. High-order Equivariant Flow Matching for Density Functional Theory Hamiltonian Prediction (NeurIPS 2025)
    Seongsu Kim, Nayoung Kim, Dongwoo Kim, and Sungsoo Ahn
    [arxiv / code]
  15. Self-Training Large Language Models with Confident Reasoning (EMNLP 2025)
    Hyosoon Jang, Yunhui Jang, Sungjae Lee, Jungseul Ok, and Sungsoo Ahn
    [arxiv]
  16. MT-Mol: Multi Agent System with Tool-based Reasoning for Molecular Optimization (EMNLP 2025)
    Hyomin Kim, Yunhui Jang, and Sungsoo Ahn
    [arxiv]
  17. Improving Chemical Understanding of LLMs via SMILES Parsing (EMNLP 2025)
    Yunhui Jang, Jaehyung Kim, and Sungsoo Ahn
    [arxiv / code]
  18. Enhancing LLM Agent Safety via Causal Influence Prompting (ACL 2025)
    Dongyoon Hahm, Woogyeol Jin, June Suk Choi, Sungsoo Ahn, and Kimin Lee
    [arxiv / code]
  19. Structural Reasoning Improves Molecular Understanding of LLM (ACL 2025)
    Yunhui Jang, Jaehyung Kim, and Sungsoo Ahn
    [arxiv / code]
  20. Generative Flows on Synthetic Pathway for Drug Design (ICLR 2025)
    Seonghwan Seo, Minsu Kim, Tony Shen, Martin Ester, Jinkyoo Park, Sungsoo Ahn, and Woo Youn Kim
    [arxiv]
  21. MOFFlow: Flow Matching for Structure Prediction of Metal-Organic Frameworks (ICLR 2025)
    Nayoung Kim, Seongsu Kim, Minsu Kim, Jinkyoo Park, and Sungsoo Ahn
    [arxiv / code]
  22. Transition Path Sampling with Improved Off-Policy Training of Diffusion Path Samplers (ICLR 2025)
    Kiyoung Seong, Seonghyun Park, Seonghwan Kim, Woo Youn Kim, and Sungsoo Ahn
    [arxiv / code / project]
  23. ReBind: Enhancing Ground-state Molecular Conformation Prediction via Force-Based Graph Rewiring (ICLR 2025)
    Taewon Kim*, Hyunjin Seo*, Sungsoo Ahn, and Eunho Yang
    [arxiv / code]
  24. Adaptive Teachers for Amortized Samplers (ICLR 2025)
    Minsu Kim, Sanghyeok Choi, Taeyoung Yun, Emmanuel Bengio, Leo Feng, Jarrid Rector-Brooks, Sungsoo Ahn, Jinkyoo Park, Nikolay Malkin, and Yoshua Bengio
    [arxiv]
  25. Decoupled Sequence and Structure Generation for Realistic Antibody Design (TMLR 2024)
    Nayoung Kim, Minsu Kim, Sungsoo Ahn, and Jinkyoo Park
    [paper / arxiv / code]
  26. Can LLMs Generate Diverse Molecules? Towards Alignment with Structural Diversity
    Hyosoon Jang, Yunhui Jang, Jaehyung Kim, and Sungsoo Ahn
    [arxiv]
  27. Iterated Energy-based Flow Matching for Sampling from Boltzmann Densities
    Dongyeop Woo and Sungsoo Ahn
    [arxiv]
  28. Non-backtracking Graph Neural Networks (TMLR 2024)
    Seonghyun Park*, Narae Ryu*, Gahee Kim, Dongyeop Woo, Se-Young Yun†, and Sungsoo Ahn†
    [paper / arxiv / code]
  29. Pessimistic Backward Policy for GFlowNets (NeurIPS 2024)
    Hyosoon Jang, Yunhui Jang, Minsu Kim, Jinkyoo Park, and Sungsoo Ahn
    [paper / arxiv / code]
  30. Hybrid Neural Representation for Spherical Data (ICML 2024)
    Hyomin Kim, Yunhui Jang, Jaeho Lee, and Sungsoo Ahn
    [paper / arxiv]
  31. Gaussian Plane-Wave Neural Operator for Electron Density Estimation (ICML 2024)
    Seongsu Kim and Sungsoo Ahn
    [paper / arxiv / code]
  32. Improving Robustness to Multiple Spurious Correlations by Multi-Objective Optimization (ICML 2024)
    Nayeong Kim, Juwon Kang, Sungsoo Ahn, Jungseul Ok, and Suha Kwak
    [paper / code]
  33. Enhancing Sample Efficiency in Black-box Combinatorial Optimization via Symmetric Replay Training (ICML 2024)
    Hyeonah Kim, Minsu Kim, Sungsoo Ahn, and Jinkyoo Park
    [arxiv / code]
  34. Tackling Complex Conditions in Unsupervised Combinatorial Optimization (ICML 2024)
    Fanchen Bu, Hyeonsoo Jo, Soo Yong Lee, Sungsoo Ahn, and Kijung Shin
    [paper / arxiv / code]
  35. Breadth-First Exploration in Adaptive Grid-based Reinforcement Learning (ICML 2024)
    Youngsik Yoon, Gangbok Lee, Sungsoo Ahn, and Jungseul Ok
    [paper / code / project]
  36. Holistic Molecular Representation Learning via Multi-view Fragmentation (TMLR 2024)
    Seojin Kim, Jaehyun Nam, Junsu Kim, Hankook Lee, Sungsoo Ahn, and Jinwoo Shin
    [paper / code]
  37. EPIC: Graph Augmentation with Edit Path Interpolation via Learnable Cost (IJCAI 2024)
    Jaeseung Heo, Seungbeom Lee, Sungsoo Ahn, and Dongwoo Kim
    [paper / arxiv]
  38. Learning Energy Decompositions for Partial Inference in GFlowNets (ICLR 2024)
    Hyosoon Jang, Minsu Kim, and Sungsoo Ahn
    [paper / arxiv / code]
  39. A Simple and Scalable Representation for Graph Generation (ICLR 2024)
    Yunhui Jang, Seul Lee, and Sungsoo Ahn
    [paper / arxiv / code]
  40. Graph Generation with K^2 Trees (ICLR 2024)
    Yunhui Jang, Dongwoo Kim, and Sungsoo Ahn
    [paper / arxiv / code]
  41. Local Search GFlowNets (ICLR 2024)
    Minsu Kim, Taeyoung Yun, Emmanuel Bengio, Dinghuai Zhang, Yoshua Bengio, Sungsoo Ahn, and Jinkyoo Park
    [paper / arxiv / code]
  42. Multi-resolution Spectral Coherence for Graph Generation with Score-based Diffusion (NeurIPS 2023)
    Hyuna Cho, Minjae Jeong, Sooyeon Jeon, Sungsoo Ahn, and Won Hwa Kim
    [paper]
  43. Diffusion Probabilistic Models for Structured Node Classification (NeurIPS 2023)
    Hyosoon Jang, Seonghyun Park, Sangwoo Mo, and Sungsoo Ahn
    [paper / arxiv / code]
  44. Bootstrapped Training of Score-Conditioned Generator for Offline Design of Biological Sequences (NeurIPS 2023)
    Minsu Kim, Federico Berto, Sungsoo Ahn, and Jinkyoo Park
    [paper / arxiv / code]
  45. A Closer Look at the Intervention Procedure of Concept Bottleneck Models (ICML 2023)
    Sungbin Shin, Yohan Jo, Sungsoo Ahn, and Namhoon Lee
    [paper / arxiv / code]
  46. Imitating Graph-Based Planning with Goal-Conditioned Policies (ICLR 2023)
    Junsu Kim, Younggyo Seo, Sungsoo Ahn, Kyunghwan Son, and Jinwoo Shin
    [paper / arxiv / code]
  47. Learning Debiased Classifier with Biased Committee (NeurIPS 2022)
    Nayeong Kim, Sehyun Hwang, Sungsoo Ahn, Jaesik Park, and Suha Kwak
    [paper / arxiv / code]
  48. Disentangling Sources of Risk for Distributional Multi-Agent Reinforcement Learning (ICML 2022)
    Kyunghwan Son, Junsu Kim, Sungsoo Ahn, Roben Delos Reyes, Yung Yi, and Jinwoo Shin
    [paper]
  49. What Makes Better Augmentation Strategies? Augment Difficult but Not Too Different (ICLR 2022)
    Jaehyung Kim, Dongyeop Kang, Sungsoo Ahn, and Jinwoo Shin
    [paper / code]
  50. Spanning Tree-based Graph Generation for Molecules (ICLR 2022)
    Sungsoo Ahn, Binghong Chen, Tianzhe Wang, and Le Song
    [paper]
  51. RoMA: Robust Model Adaptation for Offline Model-based Optimization (NeurIPS 2021)
    Sihyun Yu, Sungsoo Ahn, Le Song, and Jinwoo Shin
    [arxiv / code]
  52. Self-Improved Retrosynthetic Planning (ICML 2021)
    Junsu Kim, Sungsoo Ahn, Hankook Lee, and Jinwoo Shin
    [paper / arxiv / code]
  53. RetCL: A Selection-based Approach for Retrosynthesis via Contrastive Learning (IJCAI 2021)
    Hankook Lee, Sungsoo Ahn, Seung-Woo Seo, You Young Song, Eunho Yang, Sung Ju Hwang, and Jinwoo Shin
    [paper / arxiv / code]
  54. Layer-adaptive sparsity for the Magnitude-based Pruning (ICLR 2021)
    Jaeho Lee, Sejun Park, Sangwoo Mo, Sungsoo Ahn, and Jinwoo Shin
    [paper / arxiv / code]
  55. Learning from Failure: Training Debiased Classifier from Biased Classifier (NeurIPS 2020)
    Junhyun Nam, Hyuntak Cha, Sungsoo Ahn, Jaeho Lee, and Jinwoo Shin
    [paper / arxiv / code]
  56. Guiding Deep Molecular Optimization with Genetic Exploration (NeurIPS 2020)
    Sungsoo Ahn, Junsu Kim, Hankook Lee, and Jinwoo Shin
    [paper / arxiv / code]
  57. Learning What to Defer for Maximum Independent Sets (ICML 2020)
    Sungsoo Ahn, Younggyo Seo, and Jinwoo Shin
    [paper / arxiv / code]
  58. Variational Information Distillation for Knowledge Transfer (CVPR 2019)
    Sungsoo Ahn, Shell Hu, Andreas Damianou, Neil Lawrence, and Zhenwen Dai
    [paper / arxiv / code]
  59. Bucket-Renormalization for Approximate Inference (JSTAT 2019)
    Sungsoo Ahn, Michael Chertkov, Adrian Weller, and Jinwoo Shin
    [paper / arxiv / code]
  60. Bucket-Renormalization for Approximate Inference (ICML 2018)
    Sungsoo Ahn, Michael Chertkov, Adrian Weller, and Jinwoo Shin
    [paper / arxiv / code]
  61. Gauged Mini-Bucket Elimination for Approximate Inference (AISTATS 2018)
    Sungsoo Ahn, Michael Chertkov, Jinwoo Shin, and Adrian Weller
    [paper / arxiv / code]
  62. Gauging Variational Inference (JSTAT 2019)
    Sungsoo Ahn, Michael Chertkov, and Jinwoo Shin
    [paper / arxiv]
  63. Gauging Variational Inference (NeurIPS 2019)
    Sungsoo Ahn, Michael Chertkov, and Jinwoo Shin
    [paper / arxiv]
  64. Maximum Weight Matching using Odd-sized Cycles: Max-Product Belief Propagation and Half-Integrality (IEEE TIT 2018)
    Sungsoo Ahn, Michael Chertkov, Andrew E. Gelfand, Sejun Park, and Jinwoo Shin
    [paper]
  65. Synthesis of MCMC and Belief Propagation (NeurIPS 2016)
    Sungsoo Ahn, Michael Chertkov, and Jinwoo Shin
    [paper / arxiv]
  66. Minimum Weight Perfect Matching via Blossom Belief Propagation (NeurIPS 2015)
    Sungsoo Ahn, Sejun Park, Michael Chertkov, and Jinwoo Shin
    [paper / arxiv]