Gabriel Loaiza-Ganem

AI Researcher

Conference/journal papers, and under review preprints

  1. On Convolutions, Intrinsic Dimension, and Diffusion Models
    KK Leung, R Hosseinzadeh, G Loaiza-Ganem
    arXiv preprint 2025
  2. Textual Bayes: Quantifying Uncertainty in LLM-Based Systems
    BL Ross*, N Vouitsis*, AA Ghomi, R Hosseinzadeh, J Xin, Z Liu, Y Sui, S Hui, KK Leung, G Loaiza-Ganem, JC Cresswell
    arXiv preprint 2025
  3. A Geometric Framework for Understanding Memorization in Generative Models
    BL Ross, H Kamkari, T Wu, R Hosseinzadeh, Z Liu, G Stein, JC Cresswell, G Loaiza-Ganem
    ICLR 2025 (spotlight) [code]
  4. CaloChallenge 2022: A Community Challenge for Fast Calorimeter Simulation
    C Krause et al.
    arXiv preprint 2024
  5. A Geometric View of Data Complexity: Efficient Local Intrinsic Dimension Estimation with Diffusion Models
    H Kamkari, BL Ross, R Hosseinzadeh, JC Cresswell, G Loaiza-Ganem
    NeurIPS 2024 (spotlight) [code]
  6. Deep Generative Models through the Lens of the Manifold Hypothesis: A Survey and New Connections
    G Loaiza-Ganem, BL Ross, R Hosseinzadeh, AL Caterini, JC Cresswell
    TMLR 2024 (survey certification, expert certification) [code]
  7. A Geometric Explanation of the Likelihood OOD Detection Paradox
    H Kamkari, BL Ross, JC Cresswell, AL Caterini, RG Krishnan, G Loaiza-Ganem
    ICML 2024 [code]
  8. Data-Efficient Multimodal Fusion on a Single GPU
    N Vouitsis*, Z Liu*, SK Gorti*, V Villecroze, JC Cresswell, G Yu, G Loaiza-Ganem, M Volkovs
    CVPR 2024 (highlight) [code]
  9. Neural Implicit Manifold Learning for Topology-Aware Density Estimation
    BL Ross, G Loaiza-Ganem, AL Caterini, JC Cresswell
    TMLR 2023 (expert certification) [code]
  10. Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
    G Stein, JC Cresswell, R Hosseinzadeh, Y Sui, BL Ross, V Villecroze, Z Liu, AL Caterini*, JET Taylor*, G Loaiza-Ganem*
    NeurIPS 2023 [code]
  11. TR0N: Translator Networks for 0-Shot Plug-and-Play Conditional Generation
    Z Liu*, N Vouitsis*, SK Gorti, J Ba, G Loaiza-Ganem
    ICML 2023 [code][demo]
  12. Verifying the Union of Manifolds Hypothesis for Image Data
    BCA Brown, AL Caterini, BL Ross, JC Cresswell, G Loaiza-Ganem
    ICLR 2023 [code]
  13. Diagnosing and Fixing Manifolds Overfitting in Deep Generative Models
    G Loaiza-Ganem, BL Ross, JC Cresswell, AL Caterini
    TMLR 2022 (expert certification) [code]
  14. Bayesian Nonparametrics for Offline Skill Discovery
    V Villecroze, HJ Braviner, P Naderian, CJ Maddison, G Loaiza-Ganem
    ICML 2022 [code]
  15. Rectangular Flows for Manifold Learning
    AL Caterini*, G Loaiza-Ganem*, G Pleiss, JP Cunningham
    NeurIPS 2021 [code]
  16. C-Learning: Horizon-Aware Cumulative Accessibility Estimation
    P Naderian, G Loaiza-Ganem, HJ Braviner, AL Caterini, JC Cresswell, T Li, A Garg
    ICLR 2021 [code]
  17. Invertible Gaussian Reparameterization Trick: Revisiting the Gumbel-Softmax
    A Potapczynski, G Loaiza-Ganem, JP Cunningham
    NeurIPS 2020 [code]
  18. The Continuous Categorical: A Novel Simplex-Valued Exponential Family
    E Gordon-Rodriguez, G Loaiza-Ganem, JP Cunningham
    ICML 2020 [code]
  19. The continuous Bernoulli: fixing a pervasive error in variational autoencoders
    G Loaiza-Ganem, JP Cunningham
    NeurIPS 2019 [code]
  20. Deep Random Splines for Point Process Intensity Estimation of Neural Population Data
    G Loaiza-Ganem, SM Perkins, KE Schroeder, MM Churchland, JP Cunningham
    NeurIPS 2019 [code]
  21. Maximum Entropy Flow Networks
    G Loaiza-Ganem*, Y Gao*, JP Cunningham
    ICLR 2017 [code]



Workshop papers, abstracts, and arXiv-only preprints

  1. Last Layer Empirical Bayes
    V Villecroze, Y Wang, G Loaiza-Ganem
    ICLR 2025 ICBINB workshop [code]
  2. Deep Ensembles Secretly Perform Empirical Bayes
    G Loaiza-Ganem, V Villecroze, Y Wang
    arXiv preprint 2025
  3. Inconsistencies In Consistency Models: Better ODE Solving Does Not Imply Better Samples
    N Vouitsis, R Hosseinzadeh, BL Ross, V Villecroze, SK Gorti, JC Cresswell, G Loaiza-Ganem
    NeurIPS 2024 ATTRIB workshop [code]
  4. Inconsistencies In Consistency Models: Better ODE Solving Does Not Imply Better Samples
    N Vouitsis, R Hosseinzadeh, BL Ross, V Villecroze, SK Gorti, JC Cresswell, G Loaiza-Ganem
    NeurIPS 2024 FITML workshop [code]
  5. A Geometric View of Data Complexity: Efficient Local Intrinsic Dimension Estimation with Diffusion Models
    H Kamkari, BL Ross, R Hosseinzadeh, JC Cresswell, G Loaiza-Ganem
    ICML 2024 SPIGM workshop (oral) [code]
  6. Differentiable Local Intrinsic Dimension Estimation with Diffusion Models
    H Kamkari, BL Ross, R Hosseinzadeh, JC Cresswell, G Loaiza-Ganem
    ICML 2024 DAE workshop (oral) [code]
  7. Scalable Local Intrinsic Dimension Estimation with Diffusion Models
    H Kamkari, BL Ross, R Hosseinzadeh, JC Cresswell, G Loaiza-Ganem
    ICML 2024 GRaM workshop [code]
  8. A Geometric Framework for Understanding Memorization in Generative Models
    BL Ross, H Kamkari, Z Liu, T Wu, G Stein, G Loaiza-Ganem, JC Cresswell
    ICML 2024 GRaM workshop [code]
  9. A Geometric Framework for Understanding Memorization in Generative Models
    BL Ross, H Kamkari, Z Liu, T Wu, G Stein, G Loaiza-Ganem, JC Cresswell
    ICML 2024 NextGenAISafety workshop [code]
  10. A Geometric Explanation of the Likelihood Out-of-Distribution Detection Paradox
    H Kamkari, BL Ross, JC Cresswell, AL Caterini, RG Krishnan, G Loaiza-Ganem
    ICLR 2024 GenAI4DM workshop [code]
  11. Denoising Deep Generative Models
    G Loaiza-Ganem, BL Ross, L Wu, JP Cunningham, JC Cresswell, AL Caterini
    NeurIPS 2022 ICBINB workshop (spotlight) [code]
  12. The Union of Manifolds Hypothesis
    BCA Brown, AL Caterini, BL Ross, JC Cresswell, G Loaiza-Ganem
    NeurIPS 2022 SGNR workshop [code]
  13. Relating Regularization and Generalization through the Intrinsic Dimension of Activations
    BCA Brown, J Juravsky, AL Caterini, G Loaiza-Ganem
    NeurIPS 2022 OPT workshop
  14. Relating Regularization and Generalization through the Intrinsic Dimension of Activations
    BCA Brown, J Juravsky, AL Caterini, G Loaiza-Ganem
    NeurIPS 2022 HITY workshop
  15. CaloMan: Fast generation of calorimeter showers with density estimation on learned manifolds
    JC Cresswell, BL Ross, G Loaiza-Ganem, H Reyes-Gonzalez, M Letizia, AL Caterini
    NeurIPS 2022 ML4PS workshop [code]
  16. CaloMan: Fast generation of calorimeter showers with density estimation on learned manifolds
    JC Cresswell, BL Ross, G Loaiza-Ganem, H Reyes-Gonzalez, M Letizia, AL Caterini
    ML4Jets 2022 [code]
  17. On a novel probability distribution for zero-laden compositional data
    E Gordon-Rodriguez, G Loaiza-Ganem, JP Cunningham
    CoDaWork 2022 (oral)
  18. Neural Implicit Manifold Learning for Topology-Aware Generative Modelling
    BL Ross, G Loaiza-Ganem, AL Caterini, JC Cresswell
    ICML 2022 TAG workshop [code]
  19. On the Normalizing Constant of the Continuous Categorical Distribution
    E Gordon-Rodriguez*, G Loaiza-Ganem*, A Potapczynski, JP Cunningham
    arXiv preprint 2022 [code]
  20. Entropic Issues in Likelihood-Based OOD Detection
    AL Caterini, G Loaiza-Ganem
    NeurIPS 2021 ICBINB worshop (spotlight)
  21. Rectangular Flows for Manifold Learning
    AL Caterini*, G Loaiza-Ganem*, G Pleiss, JP Cunningham
    ICML 2021 INNF workshop (spotlight) [code]
  22. C-Learning: Horizon-Aware Cumulative Accessibility Estimation
    P Naderian, G Loaiza-Ganem, HJ Braviner, AL Caterini, JC Cresswell, T Li, A Garg
    NeurIPS 2020 DRL workshop [code]
  23. Uses and Abuses of the Cross-Entropy Loss: Case Studies in Modern Deep Learning
    E Gordon-Rodriguez, G Loaiza-Ganem, G Pleiss, JP Cunningham
    NeurIPS 2020 ICBINB workshop (oral) [code]
  24. Deep Random Splines for Point Process Intensity Estimation
    G Loaiza-Ganem, JP Cunningham
    ICLR 2019 DeepGenStruct workshop [code]
  25. Latent structure in populations of truly continuous spike trains: inference with deep random splines
    G Loaiza-Ganem, JP Cunningham
    Cosyne 2019