Karsten Kreis  

 



  ![](/sites/default/files/person/karsten_kreis_cropped_raw.jpg)

  

 Karsten Kreis is a Principal Research Scientist at NVIDIA Research focusing on generative AI.

Karsten's research interests span both the development of foundational generative AI algorithms and their application across scientific and creative domains. Recently, he has been focusing on generative learning for molecular modeling and is leading efforts in generative modeling for protein design.

Karsten is broadly excited by the intersection of generative AI and the natural sciences. In addition to scientific applications, he has also worked extensively on generative models for content creation and digital artistry—including image, video, and 3D generation.

Karsten was trained as a physicist, completing his master’s thesis in quantum information theory and his Ph.D. in computational and statistical physics, where he developed multiscale models and sampling algorithms for molecular dynamics simulations. Prior to joining NVIDIA, he worked on deep generative learning at D-Wave Systems and co-founded Variational AI, a startup leveraging generative modeling for drug discovery.

Please visit Karsten's personal website for a complete list of his publications and more information: <https://karstenkreis.github.io/>.



   Research Area(s)

[Artificial Intelligence and Machine Learning ](/research-area/machine-learning-artificial-intelligence)

[Generative AI](/research-area/generative-ai)

 

 

  

 Main Field of Interest

[Generative AI](/research-area/generative-ai)

 

  

 Google Scholar

[https://scholar.google.com/citations?user=rFd-DiAAAAAJ&amp;hl](https://scholar.google.com/citations?user=rFd-DiAAAAAJ&hl)

 

  

 

 

 



 ### Publications

 

### 2026 

[La-Proteina: Atomistic Protein Generation via Partially Latent Flow Matching](/publication/2026-01_la-proteina-atomistic-protein-generation-partially-latent-flow-matching)

[Tomas Geffner](/person/tomas-geffner), [Kieran Didi](/person/kieran-didi), Zhonglin Cao, Danny Reidenbach, Zuobai Zhang, Christian Dallago, Emine Kucukbenli, [Karsten Kreis](/person/karsten-kreis), [Arash Vahdat](/person/arash-vahdat)



[International Conference on Learning Representations (ICLR) 2026](https://arxiv.org/abs/2507.09466)









[Proteina-Complexa: Scaling Atomistic Protein Binder Design with Generative Pretraining and Test-Time Compute](/publication/2026-01_proteina-complexa-scaling-atomistic-protein-binder-design-generative)

[Kieran Didi](/person/kieran-didi), Zuobai Zhang, Guoqing Zhou, Danny Reidenbach, Zhonglin Cao, Sooyoung Cha, [Tomas Geffner](/person/tomas-geffner), Christian Dallago, Jian Tang, Michael M. Bronstein, Martin Steinegger, Emine Kucukbenli, [Arash Vahdat](/person/arash-vahdat), [Karsten Kreis](/person/karsten-kreis)



[International Conference on Learning Representations (ICLR) 2026 (Oral)](https://arxiv.org/abs/2603.27950)









[Demystifying Data-Driven Probabilistic Medium-Range Weather Forecasting](/publication/2026-01_demystifying-data-driven-probabilistic-medium-range-weather-forecasting)

[Jean Kossaifi](/person/jean-kossaifi), [Nikola Kovachki](/person/nikola-kovachki), [Morteza Mardani](/person/morteza-mardani), [Daniel Leibovici](/person/daniel-leibovici), Suman Ravuri, Ira Shokar, Edoardo Calvello, Mohammad Shoaib Abbas, Peter Harrington, Ashay Subramaniam, [Noah Brenowitz](/person/noah-brenowitz), [Boris Bonev](/person/boris-bonev), [Wonmin Byeon](/person/wonmin-byeon), [Karsten Kreis](/person/karsten-kreis), [Dale Durran](/person/dale-durran), [Arash Vahdat](/person/arash-vahdat), [Mike Pritchard](/person/mike-pritchard), [Jan Kautz](/person/jan-kautz)













[Exploring Synthesizable Chemical Space with Iterative Pathway Refinements](/publication/2026-01_exploring-synthesizable-chemical-space-iterative-pathway-refinements)

Seul Lee, [Karsten Kreis](/person/karsten-kreis), Srimukh Prasad Veccham, Meng Liu, Danny Reidenbach, Saee Paliwal, Weili Nie, [Arash Vahdat](/person/arash-vahdat)



[International Conference on Learning Representations (ICLR) 2026 (Oral)](https://arxiv.org/abs/2509.16084)









### 2025 

[Align Your Flow: Scaling Continuous-Time Flow Map Distillation](/publication/2025-12_align-your-flow-scaling-continuous-time-flow-map-distillation)

Amirmojtaba Sabour, [Sanja Fidler](/person/sanja-fidler), [Karsten Kreis](/person/karsten-kreis)



[Neural Information Processing Systems (NeurIPS) 2025](https://arxiv.org/abs/2506.14603)









[Elucidated Rolling Diffusion Models for Probabilistic Weather Forecasting](/index.php/publication/2025-12_elucidated-rolling-diffusion-models-probabilistic-weather-forecasting)

Salva Rühling Cachay, [Miika Aittala](/index.php/person/miika-aittala), [Karsten Kreis](/index.php/person/karsten-kreis), [Noah Brenowitz](/index.php/person/noah-brenowitz), [Arash Vahdat](/index.php/person/arash-vahdat), [Morteza Mardani](/index.php/person/morteza-mardani), Rose Yu



[Neural Information Processing Systems (NeurIPS) 2025](https://arxiv.org/abs/2506.20024)









[Efficient Molecular Conformer Generation with SO(3)-Averaged Flow Matching and Reflow](/publication/2025-07_efficient-molecular-conformer-generation-so3-averaged-flow-matching-and-reflow)

Zhonglin Cao, Mario Geiger, Allan Dos Santos Costa, Danny Reidenbach, [Karsten Kreis](/person/karsten-kreis), [Tomas Geffner](/person/tomas-geffner), Franco Pellegrini, Guoqing Zhou, Emine Kucukbenli



[International Conference on Machine Learning (ICML) 2025](https://arxiv.org/abs/2507.09785)









[GenMol: A Drug Discovery Generalist with Discrete Diffusion](/publication/2025-07_genmol-drug-discovery-generalist-discrete-diffusion)

Seul Lee, [Karsten Kreis](/person/karsten-kreis), Srimukh Prasad Veccham, Meng Liu, Danny Reidenbach, Yuxing Peng, Saee Paliwal, Weili Nie, [Arash Vahdat](/person/arash-vahdat)



[International Conference on Machine Learning (ICML) 2025](https://arxiv.org/abs/2501.06158)









[Score-based Diffusion Models in Function Space](/index.php/publication/2025-07_score-based-diffusion-models-function-space)

Jae Hyun Lim, [Nikola Kovachki](/index.php/person/nikola-kovachki), Ricardo Baptista, Christopher Beckham, Kamyar Azizzadenesheli, [Jean Kossaifi](/index.php/person/jean-kossaifi), Vikram Voleti, Jiaming Song, [Karsten Kreis](/index.php/person/karsten-kreis), [Jan Kautz](/index.php/person/jan-kautz), Christopher Pal, [Arash Vahdat](/index.php/person/arash-vahdat), Anima Anandkumar



[Journal of Machine Learning Research (JMLR) 2025](https://arxiv.org/abs/2302.07400)









[Multi-student Diffusion Distillation for Better One-step Generators](/publication/2025-03_multi-student-diffusion-distillation-better-one-step-generators)

Yanke Song, Jonathan Lorraine, Weili Nie, [Karsten Kreis](/person/karsten-kreis), James Lucas 



[Arxiv](https://arxiv.org/pdf/2410.23274)









[ProtComposer: Compositional Protein Structure Generation with 3D Ellipsoids](/publication/2025-01_protcomposer-compositional-protein-structure-generation-3d-ellipsoids)

Hannes Stark, Bowen Jing, [Tomas Geffner](/person/tomas-geffner), Jason Yim, Tommi Jaakkola, [Arash Vahdat](/person/arash-vahdat), [Karsten Kreis](/person/karsten-kreis)



[International Conference on Learning Representations (ICLR) 2025 (Oral)](https://arxiv.org/abs/2503.05025)









[Truncated Consistency Models](/publication/2025-01_truncated-consistency-models)

Sangyun Lee, Yilun Xu, [Tomas Geffner](/person/tomas-geffner), Giulia Fanti, [Karsten Kreis](/person/karsten-kreis), [Arash Vahdat](/person/arash-vahdat), Weili Nie



[International Conference on Learning Representations (ICLR) 2025](https://arxiv.org/abs/2410.14895)









[Proteina: Scaling Flow-based Protein Structure Generative Models](/publication/2025-01_proteina-scaling-flow-based-protein-structure-generative-models)

[Tomas Geffner](/person/tomas-geffner), [Kieran Didi](/person/kieran-didi), Zuobai Zhang, Danny Reidenbach, Zhonglin Cao, Jason Yim, Mario Geiger, Christian Dallago, Emine Kucukbenli, [Arash Vahdat](/person/arash-vahdat), [Karsten Kreis](/person/karsten-kreis)



[International Conference on Learning Representations (ICLR) 2025 (Oral)](https://arxiv.org/abs/2503.00710)









[Energy-Based Diffusion Language Models for Text Generation](/publication/2025-01_energy-based-diffusion-language-models-text-generation)

Minkai Xu, [Tomas Geffner](/person/tomas-geffner), [Karsten Kreis](/person/karsten-kreis), Weili Nie, Yilun Xu, Jure Leskovec, Stefano Ermon, [Arash Vahdat](/person/arash-vahdat)



[International Conference on Learning Representations (ICLR) 2025](https://arxiv.org/abs/2410.21357)









[Directed Graph Generation with Heat Kernels](/publication/2025-01_directed-graph-generation-heat-kernels)

Marc T. Law, [Karsten Kreis](/person/karsten-kreis), [Haggai Maron](/person/haggai-maron)



[Transactions on Machine Learning Research (TMLR) 2025](https://openreview.net/forum?id=60Gi1w6hte)









### 2024 

[Aligning Target-Aware Molecule Diffusion Models with Exact Energy Optimization](/publication/2024-12_aligning-target-aware-molecule-diffusion-models-exact-energy-optimization)

Siyi Gu, Minkai Xu, Alexander Powers, Weili Nie, [Tomas Geffner](/person/tomas-geffner), [Karsten Kreis](/person/karsten-kreis), Jure Leskovec, [Arash Vahdat](/person/arash-vahdat), Stefano Ermon



[Neural Information Processing Systems (NeurIPS) 2024](https://arxiv.org/abs/2407.01648)









[Warped Diffusion: Solving Video Inverse Problems with Image Diffusion Models](/publication/2024-12_warped-diffusion-solving-video-inverse-problems-image-diffusion-models)

Giannis Daras, Weili Nie, [Karsten Kreis](/person/karsten-kreis), Alexandros G. Dimakis, [Morteza Mardani](/person/morteza-mardani), [Nikola Kovachki](/person/nikola-kovachki), [Arash Vahdat](/person/arash-vahdat)



[Neural Information Processing Systems (NeurIPS) 2024](https://arxiv.org/abs/2410.16152)









[L4GM: Large 4D Gaussian Reconstruction Model](/publication/2024-12_l4gm-large-4d-gaussian-reconstruction-model)

Jiawei Ren, Kevin Xie, Ashkan Mirzaei, Hanxue Liang, Xiaohui Zeng, [Karsten Kreis](/person/karsten-kreis), Ziwei Liu, Antonio Torralba, Sanja Fidler, Seung Wook Kim, Huan Ling



[Neural Information Processing Systems (NeurIPS) 2024](https://arxiv.org/abs/2406.10324)









[Molecule Generation with Fragment Retrieval Augmentation](/publication/2024-12_molecule-generation-fragment-retrieval-augmentation)

Seul Lee, [Karsten Kreis](/person/karsten-kreis), Srimukh Prasad Veccham, Meng Liu, Danny Reidenbach, Saee Paliwal, [Arash Vahdat](/person/arash-vahdat), Weili Nie



[Neural Information Processing Systems (NeurIPS) 2024](https://arxiv.org/abs/2411.12078)









[DisCo-Diff: Enhancing Continuous Diffusion Models with Discrete Latents](/publication/2024-07_disco-diff-enhancing-continuous-diffusion-models-discrete-latents)

Yilun Xu, Gabriele Corso, Tommi Jaakkola, [Arash Vahdat](/person/arash-vahdat), [Karsten Kreis](/person/karsten-kreis)



[International Conference on Machine Learning (ICML) 2024](https://arxiv.org/abs/2407.03300)









[Align Your Steps: Optimizing Sampling Schedules in Diffusion Models](/publication/2024-07_align-your-steps-optimizing-sampling-schedules-diffusion-models)

Amirmojtaba Sabour, Sanja Fidler, [Karsten Kreis](/person/karsten-kreis)



[International Conference on Machine Learning (ICML) 2024](https://arxiv.org/abs/2404.14507)









[Outdoor Scene Extrapolation with Hierarchical Generative Cellular Automata](/publication/2024-06_outdoor-scene-extrapolation-hierarchical-generative-cellular-automata)

Dongsu Zhang, Francis Williams, Zan Gojcic, [Karsten Kreis](/person/karsten-kreis), Sanja Fidler, Young Min Kim, Amlan Kar



[CVPR 2024 (Highlight)](https://arxiv.org/abs/2406.08292)









[Align Your Gaussians: Text-to-4D with Dynamic 3D Gaussians and Composed Diffusion Models](/publication/2024-06_align-your-gaussians-text-4d-dynamic-3d-gaussians-and-composed-diffusion-models)

Huan Ling, Seung Wook Kim, Antonio Torralba, Sanja Fidler, [Karsten Kreis](/person/karsten-kreis)



[CVPR 2024 (Highlight)](https://arxiv.org/abs/2312.13763)









[WildFusion: Learning 3D-Aware Latent Diffusion Models in View Space](/index.php/publication/2024-05_wildfusion-learning-3d-aware-latent-diffusion-models-view-space)

Katja Schwarz, Seung Wook Kim, Jun Gao, Sanja Fidler, Andreas Geiger, [Karsten Kreis](/index.php/person/karsten-kreis)



[International Conference on Learning Representations (ICLR) 2024](https://arxiv.org/abs/2311.13570)









### 2023 

[TexFusion: Synthesizing 3D Textures with Text-Guided Image Diffusion Models](/publication/2023-10_texfusion-synthesizing-3d-textures-text-guided-image-diffusion-models)

Tianshi Cao, [Karsten Kreis](/person/karsten-kreis), Sanja Fidler, Nicholas Sharp, Kangxue Yin



[IEEE/CVF International Conference on Computer Vision (ICCV) 2023 (Oral)](https://arxiv.org/abs/2310.13772)









[DreamTeacher: Pretraining Image Backbones with Deep Generative Models](/publication/2023-10_dreamteacher-pretraining-image-backbones-deep-generative-models)

Daiqing Li, Huan Ling, Amlan Kar, David Acuna, Seung Wook Kim, [Karsten Kreis](/person/karsten-kreis), Antonio Torralba, Sanja Fidler



[IEEE/CVF International Conference on Computer Vision (ICCV) 2023](https://arxiv.org/abs/2307.07487)









[Differentially Private Diffusion Models](/publication/2023-08_differentially-private-diffusion-models)

Tim Dockhorn, Tianshi Cao, [Arash Vahdat](/person/arash-vahdat), [Karsten Kreis](/person/karsten-kreis)



[Transactions on Machine Learning Research (TMLR) 2023](https://arxiv.org/abs/2210.09929)









[Align your Latents: High-Resolution Video Synthesis with Latent Diffusion Models](/publication/2023-06_align-your-latents-high-resolution-video-synthesis-latent-diffusion-models)

Andreas Blattmann, Robin Rombach, Huan Ling, Tim Dockhorn, Seung Wook Kim, Sanja Fidler, [Karsten Kreis](/person/karsten-kreis)



[IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2023](https://arxiv.org/abs/2304.08818)









[NeuralField-LDM: Scene Generation with Hierarchical Latent Diffusion Models](/publication/2023-06_neuralfield-ldm-scene-generation-hierarchical-latent-diffusion-models)

Seung Wook Kim, Bradley Brown, Kangxue Yin, [Karsten Kreis](/person/karsten-kreis), Katja Schwarz, Daiqing Li, Robin Rombach, Antonio Torralba, Sanja Fidler



[IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2023](https://arxiv.org/abs/2304.09787)









[Trace and Pace: Controllable Pedestrian Animation via Guided Trajectory Diffusion](/publication/2023-06_trace-and-pace-controllable-pedestrian-animation-guided-trajectory-diffusion)

Davis Rempe, Zhengyi Luo, Xue Bin Peng, [Ye Yuan](/person/ye-yuan), Kris Kitani, [Karsten Kreis](/person/karsten-kreis), Sanja Fidler, Or Litany



[IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2023](https://cvpr2023.thecvf.com/)









[Magic3D: High-Resolution Text-to-3D Content Creation](/index.php/publication/2023-06_magic3d-high-resolution-text-3d-content-creation)

[Chen-Hsuan Lin](/index.php/person/chen-hsuan-lin), Jun Gao, Luming Tang, Towaki Takikawa, Xiaohui Zeng, Xun Huang, [Karsten Kreis](/index.php/person/karsten-kreis), Sanja Fidler, [Ming-Yu Liu](/index.php/person/ming-yu-liu), [Tsung-Yi Lin](/index.php/person/tsung-yi-lin)



[CVPR 2023 (Highlight)](https://cvpr2023.thecvf.com/)









### 2022 

[LION: Latent Point Diffusion Models for 3D Shape Generation](/index.php/publication/2022-11_lion-latent-point-diffusion-models-3d-shape-generation)

Xiaohui Zeng, [Arash Vahdat](/index.php/person/arash-vahdat), Francis Williams, Zan Gojcic, Or Litany, Sanja Fidler, [Karsten Kreis](/index.php/person/karsten-kreis)



[Neural Information Processing Systems (NeurIPS) 2022](https://arxiv.org/abs/2210.06978)









[GENIE: Higher-Order Denoising Diffusion Solvers](/publication/2022-11_genie-higher-order-denoising-diffusion-solvers)

Tim Dockhorn, [Arash Vahdat](/person/arash-vahdat), [Karsten Kreis](/person/karsten-kreis)



[Neural Information Processing Systems (NeurIPS) 2022](https://arxiv.org/abs/2210.05475)









[Polymorphic-GAN: Generating Aligned Samples across Multiple Domains with Learned Morph Maps](/publication/2022-06_polymorphic-gan-generating-aligned-samples-across-multiple-domains-learned)

Seung Wook Kim, [Karsten Kreis](/person/karsten-kreis), Daiqing Li, Antonio Torralba, Sanja Fidler



[Conference on Computer Vision and Pattern Recognition (CVPR) 2022 (Oral)](https://arxiv.org/abs/2206.02903)









[BigDatasetGAN: Synthesizing ImageNet with Pixel-wise Annotations](/publication/2022-06_bigdatasetgan-synthesizing-imagenet-pixel-wise-annotations)

Daiqing Li, Huan Ling, Seung Wook Kim, [Karsten Kreis](/person/karsten-kreis), Adela Barriuso, Sanja Fidler, Antonio Torralba



[IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2022](https://arxiv.org/abs/2201.04684)









[ Tackling the Generative Learning Trilemma with Denoising Diffusion GANs](/publication/2022-03_tackling-generative-learning-trilemma-denoising-diffusion-gans-0)

Zhisheng Xiao, [Karsten Kreis](/person/karsten-kreis), [Arash Vahdat](/person/arash-vahdat)



[International Conference on Learning Representations (ICLR) 2022 (Spotlight)](https://arxiv.org/abs/2112.07804)









[Score-Based Generative Modeling with Critically-Damped Langevin Diffusion](/publication/2022-03_score-based-generative-modeling-critically-damped-langevin-diffusion)

Tim Dockhorn, [Arash Vahdat](/person/arash-vahdat), [Karsten Kreis](/person/karsten-kreis)



[International Conference on Learning Representations (ICLR) 2022 (Spotlight)](https://arxiv.org/abs/2112.07068)









### 2021 

[ATISS: Autoregressive Transformers for Indoor Scene Synthesis](/publication/2021-12_atiss-autoregressive-transformers-indoor-scene-synthesis)

Despoina Paschalidou, Amlan Kar, Maria Shugrina, [Karsten Kreis](/person/karsten-kreis), Andreas Geiger, Sanja Fidler



[Neural Information Processing Systems (NeurIPS) 2021](https://arxiv.org/abs/2110.03675)









[EditGAN: High-Precision Semantic Image Editing](/publication/2021-12_editgan-high-precision-semantic-image-editing)

Huan Ling, [Karsten Kreis](/person/karsten-kreis), Daiqing Li, Seung Wook Kim, Antonio Torralba, Sanja Fidler



[Neural Information Processing Systems (NeurIPS) 2021](https://arxiv.org/abs/2111.03186)









[Don’t Generate Me: Training Differentially Private Generative Models with Sinkhorn Divergence](/publication/2021-11_don-t-generate-me-training-differentially-private-generative-models-sinkhorn)

Tianshi Cao, Alex Bie, [Arash Vahdat](/person/arash-vahdat), Sanja Fidler, [Karsten Kreis](/person/karsten-kreis)



[Neural Information Processing Systems (NeurIPS) 2021](https://arxiv.org/abs/2111.01177)









[Score-based Generative Modeling in Latent Space](/index.php/publication/2021-11_score-based-generative-modeling-latent-space)

[Arash Vahdat](/index.php/person/arash-vahdat), [Karsten Kreis](/index.php/person/karsten-kreis), [Jan Kautz](/index.php/person/jan-kautz)



[Neural Information Processing Systems (NeurIPS) 2021](https://arxiv.org/abs/2106.05931)









[Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D Shapes](/nglod)

Towaki Takikawa, Joey Litalien, Kangxue Yin, [Karsten Kreis](/person/karsten-kreis), [Charles Loop](/person/charles-loop), Derek Nowrouzezahrai, Alec Jacobson, Morgan McGuire, Sanja Fidler



[IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2021](https://cvpr2021.thecvf.com/)









[Semantic Segmentation with Generative Models: Semi-Supervised Learning and Strong Out-of-Domain Generalization](/publication/2021-06_semantic-segmentation-generative-models-semi-supervised-learning-and-strong-out)

Daiqing Li, Junlin Yang, [Karsten Kreis](/person/karsten-kreis), Antonio Torralba, Sanja Fidler



[IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2021](https://arxiv.org/abs/2104.05833)









[VAEBM: A Symbiosis between Variational Autoencoders and Energy-based Models](/publication/2021-06_vaebm-symbiosis-between-variational-autoencoders-and-energy-based-models)

Zhisheng Xiao, [Karsten Kreis](/person/karsten-kreis), [Jan Kautz](/person/jan-kautz), [Arash Vahdat](/person/arash-vahdat)



[International Conference on Learning Representations (ICLR) 2021 (Spotlight)](https://arxiv.org/abs/2010.00654)









### 2020 

[Variational Amodal Object Completion](/publication/2020-12_variational-amodal-object-completion)

Huan Ling, David Acuna, [Karsten Kreis](/person/karsten-kreis), Seung Wook Kim, Sanja Fidler



[Neural Information Processing Systems (NeurIPS) 2020](https://papers.nips.cc/paper/2020/hash/bacadc62d6e67d7897cef027fa2d416c-Abstract.html)