arxiv: 2603.21717 · v3 · submitted 2026-03-23 · 💻 cs.LG

Recognition: no theorem link

Uncertainty Quantification for Distribution-to-Distribution Flow Matching in Scientific Imaging

Dongxia Wu , Yuhui Zhang , Serena Yeung-Levy , Emma Lundberg , Emily B. Fox

Authors on Pith no claims yet

Pith reviewed 2026-05-15 01:07 UTC · model grok-4.3

classification 💻 cs.LG

keywords uncertainty quantificationflow matchingscientific imagingaleatoric uncertaintyepistemic uncertaintyout-of-distribution detectiongenerative modelsdistribution-to-distribution

0 comments

The pith

Bayesian Stochastic Flow Matching disentangles aleatoric and epistemic uncertainty in distribution-to-distribution generative models.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces Bayesian Stochastic Flow Matching as a unified framework for uncertainty quantification in generative models that map one data distribution to another. These models are used in scientific imaging to simulate cellular responses to perturbations or to translate medical scans across conditions. The approach augments deterministic flow matching with a diffusion term to handle variations across labs and devices, while a new sampling method called MCD-Antithetic combines Monte Carlo dropout with antithetic sampling to produce anomaly scores. This setup separates uncertainty due to inherent data noise from uncertainty due to limited model knowledge. A reader would care because it provides a concrete way to trust or flag model outputs when experimental conditions shift.

Core claim

The central claim is that Stochastic Flow Matching augments deterministic flows with a diffusion term to improve generalization to unseen scenarios in distribution-to-distribution tasks, and that the MCD-Antithetic Bayesian approach yields effective anomaly scores for out-of-distribution detection while disentangling aleatoric from epistemic uncertainty. Experiments on cellular imaging datasets BBBC021 and JUMP plus brain fMRI data from the Theory of Mind task demonstrate that the stochastic component enhances reliability across conditions and that the sampling method improves accountability through better anomaly detection.

What carries the argument

Bayesian Stochastic Flow Matching (BSFM), which integrates Stochastic Flow Matching (SFM) that adds a diffusion term to deterministic flows for better generalization, together with MCD-Antithetic sampling that combines Monte Carlo dropout and antithetic sampling to generate scalable anomaly scores.

If this is right

SFM improves reliability of generated cellular perturbation responses across varied experimental setups.
MCD-Antithetic produces anomaly scores that enhance detection of cases where predictions may be unreliable.
The disentangled uncertainties allow separate handling of data noise versus model limitations in medical image translation tasks.
The framework supports accountable use of distribution-to-distribution models on fMRI data under diverse conditions.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same stochastic augmentation might apply to flow matching in non-imaging domains such as molecular structure generation.
High epistemic uncertainty regions identified by the method could guide targeted data collection to reduce model ignorance.
Integration with existing calibration techniques could further tighten the separation between the two uncertainty types.

Load-bearing premise

The assumption that augmenting deterministic flows with a diffusion term improves generalization to new experimental conditions and that MCD-Antithetic sampling reliably produces effective anomaly scores.

What would settle it

A controlled comparison on held-out imaging data from new labs or devices showing no gain in generalization metrics for SFM or no improvement in out-of-distribution detection accuracy for MCD-Antithetic relative to standard deterministic flow matching baselines.

Figures

Figures reproduced from arXiv: 2603.21717 by Dongxia Wu, Emily B. Fox, Emma Lundberg, Serena Yeung-Levy, Yuhui Zhang.

**Figure 2.** Figure 2: Core components of the Bayesian Stochastic Flow Matching (BSFM) framework: (a) Stochastic flow match [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Figure 3: Examples of generated images from different methods under various unseen scenarios, compared with the [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗

**Figure 4.** Figure 4: Examples of generated images from different methods on BBBC021 under Unseen Pert. OOD scenario, [PITH_FULL_IMAGE:figures/full_fig_p016_4.png] view at source ↗

**Figure 5.** Figure 5: Examples of generated images from different methods on BBBC021 under Intensity Shift OOD scenario, [PITH_FULL_IMAGE:figures/full_fig_p016_5.png] view at source ↗

**Figure 6.** Figure 6: Examples of generated images from different methods on JUMP under Unseen Cell Lines OOD scenario, [PITH_FULL_IMAGE:figures/full_fig_p017_6.png] view at source ↗

**Figure 7.** Figure 7: Examples of generated images from different methods on JUMP under Unseen Plates OOD scenario, com [PITH_FULL_IMAGE:figures/full_fig_p017_7.png] view at source ↗

read the original abstract

Distribution-to-distribution generative models support scientific imaging tasks ranging from modeling cellular perturbation responses to translating medical images across conditions. Trustworthy generation requires both reliability (generalization across labs, devices, and experimental conditions) and accountability (detecting out-of-distribution cases where predictions may be unreliable). Uncertainty quantification (UQ) based approaches serve as promising candidates for these tasks, yet UQ for distribution-to-distribution generative models remains underexplored. We present a unified UQ framework, Bayesian Stochastic Flow Matching (BSFM), that disentangles aleatoric and epistemic uncertainty. The Stochastic Flow Matching (SFM) component augments deterministic flows with a diffusion term to improve model generalization to unseen scenarios. For UQ, we develop a scalable Bayesian approach -- MCD-Antithetic -- that combines Monte Carlo Dropout with sample-efficient antithetic sampling to produce effective anomaly scores for out-of-distribution detection. Experiments on cellular imaging (BBBC021, JUMP) and brain fMRI (Theory of Mind) across diverse scenarios show that SFM improves reliability while MCD-Antithetic enhances accountability.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

BSFM adds a diffusion term to flow matching and pairs it with antithetic dropout for UQ in imaging tasks, but the paper does not derive how the stochasticity preserves the original marginals or cleanly separates uncertainty types.

read the letter

The paper's core move is to extend flow matching with an explicit diffusion term, call the result Stochastic Flow Matching, and then wrap it in a Bayesian setup using Monte Carlo dropout with antithetic sampling to produce separate aleatoric and epistemic uncertainty estimates plus anomaly scores. They test the whole thing on cellular perturbation imaging from BBBC021 and JUMP plus one fMRI dataset, reporting better generalization across conditions and stronger out-of-distribution detection than plain flow matching baselines. That combination is new in this exact setting and the experiments use real scientific data rather than toy benchmarks, which gives the work some immediate relevance for people who need generative models that flag when they are likely to be wrong. The practical framing around reliability across labs and devices is reasonable and the datasets are appropriate for the claims. The main weakness is that the justification for the diffusion term stays at the level of assertion. There is no derivation showing that the added stochasticity leaves the learned conditional distribution unchanged in the limit or that it supplies an aleatoric component independent of the epistemic one recovered by dropout. The experiments show aggregate gains but do not include an ablation that keeps model capacity fixed and removes only the learned diffusion schedule, so it remains unclear how much of the reported improvement traces to the new term versus extra flexibility. The MCD-Antithetic procedure itself is a straightforward engineering choice and appears to work on the reported tasks, yet its sample-efficiency advantage is not demonstrated against standard dropout on identical architectures. This paper is for researchers already working on distribution-to-distribution models in biology or medical imaging who want a concrete UQ recipe they can try on their own data. It is not a foundational theoretical advance, but the application focus and the dataset results make it worth a referee's time to check whether the stochastic extension actually delivers the claimed separation and generalization benefits.

Referee Report

2 major / 2 minor

Summary. The paper proposes a unified uncertainty quantification framework called Bayesian Stochastic Flow Matching (BSFM) for distribution-to-distribution generative models in scientific imaging. It introduces Stochastic Flow Matching (SFM) by augmenting deterministic flow-matching with an explicit diffusion term to improve generalization to unseen scenarios, and develops MCD-Antithetic (Monte Carlo Dropout combined with antithetic sampling) to disentangle aleatoric and epistemic uncertainty while producing anomaly scores for out-of-distribution detection. Experiments on cellular imaging datasets (BBBC021, JUMP) and brain fMRI (Theory of Mind) across diverse conditions are reported to demonstrate gains in reliability and accountability.

Significance. If the central claims are substantiated, this would represent a meaningful advance in trustworthy generative modeling for scientific domains, supplying a practical method to separate uncertainty types in flow-based models and thereby supporting better generalization across labs/devices and reliable anomaly detection in high-stakes imaging tasks.

major comments (2)

[Stochastic Flow Matching formulation (likely §3 or §4)] The manuscript asserts that augmenting the deterministic flow-matching objective with a diffusion term (SFM) leaves the learned conditional distribution unchanged in the limit while cleanly isolating an aleatoric component; however, no derivation of the modified continuity equation or proof that the added stochasticity does not bias the velocity field is provided, which is load-bearing for the disentanglement claim.
[Experimental results (likely §5)] Experiments on BBBC021 and JUMP report aggregate improvements in reliability without an ablation that holds model capacity fixed and removes only the learned diffusion schedule; this omission prevents attribution of gains specifically to SFM rather than increased expressivity.

minor comments (2)

[Abstract] The abstract introduces 'MCD-Antithetic' and 'BSFM' without a brief parenthetical definition or reference to the defining equations, which reduces immediate readability.
[Methods and Experiments] Notation for the diffusion schedule and antithetic sampling variance reduction is not introduced with sufficient clarity before the experimental tables, making it difficult to interpret the reported anomaly scores.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their insightful comments, which have helped us improve the clarity and rigor of our work. We address each major comment in detail below and outline the revisions we plan to make.

read point-by-point responses

Referee: [Stochastic Flow Matching formulation (likely §3 or §4)] The manuscript asserts that augmenting the deterministic flow-matching objective with a diffusion term (SFM) leaves the learned conditional distribution unchanged in the limit while cleanly isolating an aleatoric component; however, no derivation of the modified continuity equation or proof that the added stochasticity does not bias the velocity field is provided, which is load-bearing for the disentanglement claim.

Authors: We thank the referee for pointing out the need for a formal derivation. The current version of the manuscript introduces the SFM formulation in Section 3 and demonstrates its empirical benefits, but we agree that including a derivation of the modified continuity equation is essential to substantiate the claim that the conditional distribution remains unchanged in the limit and that the stochastic term isolates the aleatoric uncertainty without biasing the velocity field. In the revised manuscript, we will add this derivation, showing how the stochastic process corresponds to a Fokker-Planck equation that reduces to the deterministic flow-matching case under the appropriate scaling, thereby preserving the target distribution while allowing for explicit modeling of aleatoric noise. This will directly support the disentanglement in the BSFM framework. revision: yes
Referee: [Experimental results (likely §5)] Experiments on BBBC021 and JUMP report aggregate improvements in reliability without an ablation that holds model capacity fixed and removes only the learned diffusion schedule; this omission prevents attribution of gains specifically to SFM rather than increased expressivity.

Authors: We acknowledge that the experiments as presented report overall performance gains without a capacity-controlled ablation isolating the effect of the diffusion schedule. To address this, we will include an additional ablation study in the revised version. Specifically, we will train models with identical architectures and parameter counts, comparing the deterministic flow-matching baseline against the SFM variant (with the learned diffusion schedule) on the BBBC021 and JUMP datasets. This will allow us to attribute improvements specifically to the stochastic augmentation rather than general increases in model capacity or expressivity. We expect this to further validate the SFM component. revision: yes

Circularity Check

0 steps flagged

No significant circularity; claims rest on experimental results rather than self-referential definitions

full rationale

The abstract and description introduce BSFM, SFM (augmenting flows with diffusion), and MCD-Antithetic without exhibiting any equations, fitted parameters, or continuity-equation derivations that reduce the claimed disentanglement or generalization improvement to the inputs by construction. No self-citation load-bearing steps, uniqueness theorems, or ansatz smuggling are visible. The skeptic concern about missing derivations for the modified continuity equation is a correctness or completeness issue, not a circularity reduction. The paper's central claims are therefore treated as independent of any tautological fit or renaming.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 2 invented entities

Abstract-only review limits visibility into parameters and assumptions; the framework appears to rest on standard flow matching plus new stochastic and Bayesian additions without independent evidence for the new components.

axioms (1)

domain assumption Augmenting deterministic flow matching with a diffusion term improves generalization to unseen experimental conditions
Invoked in the description of the SFM component as a means to support reliability across labs and devices.

invented entities (2)

Bayesian Stochastic Flow Matching (BSFM) no independent evidence
purpose: Unified framework disentangling aleatoric and epistemic uncertainty
Newly introduced combination of stochastic flow and Bayesian UQ
MCD-Antithetic no independent evidence
purpose: Scalable Bayesian sampling method for anomaly scores
Developed specifically for out-of-distribution detection in this setting

pith-pipeline@v0.9.0 · 5496 in / 1289 out tokens · 66307 ms · 2026-05-15T01:07:12.680827+00:00 · methodology

discussion (0)

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

Divergence is Uncertainty: A Closed-Form Posterior Covariance for Flow Matching
cs.LG 2026-05 unverdicted novelty 8.0

In flow matching, the uncertainty of the clean data given the current state is exactly the divergence of the velocity field (up to a known scalar).

Reference graph

Works this paper leans on

55 extracted references · 55 canonical work pages · cited by 1 Pith paper · 7 internal anchors

[1]

Zero-shot image anomaly detection using generative foundation models.arXiv preprint arXiv:2507.22692, 2025

Lemar Abdi, Amaan Valiuddin, Francisco Caetano, Christiaan Viviers, and Fons van der Sommen. Zero-shot image anomaly detection using generative foundation models.arXiv preprint arXiv:2507.22692, 2025. 3

work page arXiv 2025
[2]

Stochastic Interpolants: A Unifying Framework for Flows and Diffusions

Michael S Albergo, Nicholas M Boffi, and Eric Vanden-Eijnden. Stochastic interpolants: A unifying framework for flows and diffusions.arXiv preprint arXiv:2303.08797, 2023. 5

work page internal anchor Pith review Pith/arXiv arXiv 2023
[3]

Self-consistent recursive diffusion bridge for medical image translation.Medical Image Analysis, 106:103747, 2025

Fuat Arslan, Bilal Kabas, Onat Dalmaz, Muzaffer Ozbey, and Tolga C ¸ ukur. Self-consistent recursive diffusion bridge for medical image translation.Medical Image Analysis, 106:103747, 2025. 1, 3

work page 2025
[4]

Diffusion models with implicit guid- ance for medical anomaly detection

Cosmin I Bercea, Benedikt Wiestler, Daniel Rueckert, and Julia A Schnabel. Diffusion models with implicit guid- ance for medical anomaly detection. InInternational Conference on Medical Image Computing and Computer- Assisted Intervention, pages 211–220. Springer, 2024. 3

work page 2024
[5]

Shedding light on large generative networks: Estimating epistemic uncertainty in diffusion models

Lucas Berry, Axel Brando, and David Meger. Shedding light on large generative networks: Estimating epistemic uncertainty in diffusion models. InThe 40th Conference on Uncertainty in Artificial Intelligence, 2024. 3

work page 2024
[6]

Protecting scientific integrity in an age of generative ai, 2024

Wolfgang Blau, Vinton G Cerf, Juan Enriquez, Joseph S Francisco, Urs Gasser, Mary L Gray, Mark Greaves, Barbara J Grosz, Kathleen Hall Jamieson, Gerald H Haug, et al. Protecting scientific integrity in an age of generative ai, 2024. 1

work page 2024
[7]

High-content phenotypic profiling of drug response signatures across distinct cancer cells.Molecular cancer therapeutics, 9(6):1913–1926, 2010

Peter D Caie, Rebecca E Walls, Alexandra Ingleston-Orme, Sandeep Daya, Tom Houslay, Rob Eagle, Mark E Roberts, and Neil O Carragher. High-content phenotypic profiling of drug response signatures across distinct cancer cells.Molecular cancer therapeutics, 9(6):1913–1926, 2010. 7, 14

work page 1913
[8]

Estimating epistemic and aleatoric uncertainty with a single model.Advances in Neural Information Processing Systems, 37:109845–109870, 2024

Matthew Chan, Maria Molina, and Chris Metzler. Estimating epistemic and aleatoric uncertainty with a single model.Advances in Neural Information Processing Systems, 37:109845–109870, 2024. 3

work page 2024
[9]

Jump cell painting dataset: morphological impact of 136,000 chemical and genetic perturbations.BioRxiv, pages 2023–03, 2023

Srinivas Niranj Chandrasekaran, Jeanelle Ackerman, Eric Alix, D Michael Ando, John Arevalo, Melissa Ben- nion, Nicolas Boisseau, Adriana Borowa, Justin D Boyd, Laurent Brino, et al. Jump cell painting dataset: morphological impact of 136,000 chemical and genetic perturbations.BioRxiv, pages 2023–03, 2023. 7, 14

work page 2023
[10]

Projection regret: Reducing background bias for novelty detection via diffusion models.Advances in Neural Information Processing Systems, 36:19230–19245,

Sungik Choi, Hankook Lee, Honglak Lee, and Moontae Lee. Projection regret: Reducing background bias for novelty detection via diffusion models.Advances in Neural Information Processing Systems, 36:19230–19245,

work page
[11]

Diffusion schr ¨odinger bridge with applications to score-based generative modeling.Advances in neural information processing systems, 34:17695– 17709, 2021

Valentin De Bortoli, James Thornton, Jeremy Heng, and Arnaud Doucet. Diffusion schr ¨odinger bridge with applications to score-based generative modeling.Advances in neural information processing systems, 34:17695– 17709, 2021. 3 10 UQ for Dist-to-Dist Flow Matching in Scientific ImagingA PREPRINT

work page 2021
[12]

Diffusion model guided sampling with pixel-wise aleatoric uncer- tainty estimation

Michele De Vita and Vasileios Belagiannis. Diffusion model guided sampling with pixel-wise aleatoric uncer- tainty estimation. In2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), pages 3844–3854. IEEE, 2025. 3

work page 2025
[13]

Conformalized generative bayesian imaging: An uncertainty quantification framework for computational imaging.arXiv preprint arXiv:2504.07696, 2025

Canberk Ekmekci and Mujdat Cetin. Conformalized generative bayesian imaging: An uncertainty quantification framework for computational imaging.arXiv preprint arXiv:2504.07696, 2025. 3

work page arXiv 2025
[14]

Score-based diffusion models as principled priors for inverse imaging

Berthy T Feng, Jamie Smith, Michael Rubinstein, Huiwen Chang, Katherine L Bouman, and William T Free- man. Score-based diffusion models as principled priors for inverse imaging. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 10520–10531, 2023. 3

work page 2023
[15]

Towards understanding and quantifying uncertainty for text-to-image generation

Gianni Franchi, Nacim Belkhir, Dat Nguyen Trong, Guoxuan Xia, and Andrea Pilzer. Towards understanding and quantifying uncertainty for text-to-image generation. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 8062–8072, 2025. 3

work page 2025
[16]

Dropout as a bayesian approximation: Representing model uncertainty in deep learning

Yarin Gal and Zoubin Ghahramani. Dropout as a bayesian approximation: Representing model uncertainty in deep learning. Ininternational conference on machine learning, pages 1050–1059. PMLR, 2016. 6

work page 2016
[17]

Diffusion for out-of-distribution detection on road scenes and beyond

Silvio Galesso, Philipp Schr ¨oppel, Hssan Driss, and Thomas Brox. Diffusion for out-of-distribution detection on road scenes and beyond. InEuropean Conference on Computer Vision, pages 110–126. Springer, 2024. 3

work page 2024
[18]

Diffguard: Semantic mismatch-guided out- of-distribution detection using pre-trained diffusion models

Ruiyuan Gao, Chenchen Zhao, Lanqing Hong, and Qiang Xu. Diffguard: Semantic mismatch-guided out- of-distribution detection using pre-trained diffusion models. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 1579–1589, 2023. 3

work page 2023
[19]

Denoising diffusion models for out-of-distribution detection

Mark S Graham, Walter HL Pinaya, Petru-Daniel Tudosiu, Parashkev Nachev, Sebastien Ourselin, and Jorge Car- doso. Denoising diffusion models for out-of-distribution detection. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 2948–2957, 2023. 3

work page 2023
[20]

Unsupervised 3d out-of-distribution detection with latent diffusion models

Mark S Graham, Walter Hugo Lopez Pinaya, Paul Wright, Petru-Daniel Tudosiu, Yee H Mah, James T Teo, H Rolf J¨ager, David Werring, Parashkev Nachev, Sebastien Ourselin, et al. Unsupervised 3d out-of-distribution detection with latent diffusion models. InInternational Conference on Medical Image Computing and Computer- Assisted Intervention, pages 446–456. ...

work page 2023
[21]

A new monte carlo technique: antithetic variates

John Michael Hammersley and Keith William Morton. A new monte carlo technique: antithetic variates. In Mathematical proceedings of the Cambridge philosophical society, volume 52, pages 449–475. Cambridge Uni- versity Press, 1956. 6

work page 1956
[22]

Out-of-distribution detection with a single unconditional diffusion model.Ad- vances in Neural Information Processing Systems, 37:43952–43974, 2024

Alvin Heng, Harold Soh, et al. Out-of-distribution detection with a single unconditional diffusion model.Ad- vances in Neural Information Processing Systems, 37:43952–43974, 2024. 3

work page 2024
[23]

Classifier-Free Diffusion Guidance

Jonathan Ho and Tim Salimans. Classifier-free diffusion guidance.arXiv preprint arXiv:2207.12598, 2022. 4

work page internal anchor Pith review Pith/arXiv arXiv 2022
[24]

On the trustworthiness of generative foundation models: Guideline, assessment, and perspective.arXiv preprint arXiv:2502.14296, 2025

Yue Huang, Chujie Gao, Siyuan Wu, Haoran Wang, Xiangqi Wang, Yujun Zhou, Yanbo Wang, Jiayi Ye, Jiawen Shi, Qihui Zhang, et al. On the trustworthiness of generative foundation models: Guideline, assessment, and perspective.arXiv preprint arXiv:2502.14296, 2025. 1

work page arXiv 2025
[25]

Generative uncertainty in diffusion models.arXiv preprint arXiv:2502.20946, 2025

Metod Jazbec, Eliot Wong-Toi, Guoxuan Xia, Dan Zhang, Eric Nalisnick, and Stephan Mandt. Generative uncertainty in diffusion models.arXiv preprint arXiv:2502.20946, 2025. 3

work page arXiv 2025
[26]

Fbnetgen: Task-aware gnn-based fmri analysis via functional brain network generation

Xuan Kan, Hejie Cui, Joshua Lukemire, Ying Guo, and Carl Yang. Fbnetgen: Task-aware gnn-based fmri analysis via functional brain network generation. InInternational conference on medical imaging with deep learning, pages 618–637. PMLR, 2022. 1, 3

work page 2022
[27]

Unpaired image-to-image translation via neural schr\” odinger bridge.arXiv preprint arXiv:2305.15086, 2023

Beomsu Kim, Gihyun Kwon, Kwanyoung Kim, and Jong Chul Ye. Unpaired image-to-image translation via neural schr\” odinger bridge.arXiv preprint arXiv:2305.15086, 2023. 3, 7

work page arXiv 2023
[28]

Unsupervised anomaly detection using diffusion trend analysis for display inspection.arXiv preprint arXiv:2407.09578, 2024

Eunwoo Kim, Un Yang, Cheol Lae Roh, and Stefano Ermon. Unsupervised anomaly detection using diffusion trend analysis for display inspection.arXiv preprint arXiv:2407.09578, 2024. 3

work page arXiv 2024
[29]

Why normalizing flows fail to detect out-of- distribution data.Advances in neural information processing systems, 33:20578–20589, 2020

Polina Kirichenko, Pavel Izmailov, and Andrew G Wilson. Why normalizing flows fail to detect out-of- distribution data.Advances in neural information processing systems, 33:20578–20589, 2020. 8

work page 2020
[30]

Bayesdiff: Estimating pixel-wise uncertainty in diffusion via bayesian inference.arXiv preprint arXiv:2310.11142, 2023

Siqi Kou, Lei Gan, Dequan Wang, Chongxuan Li, and Zhijie Deng. Bayesdiff: Estimating pixel-wise uncertainty in diffusion via bayesian inference.arXiv preprint arXiv:2310.11142, 2023. 3, 7

work page arXiv 2023
[31]

Predicting task-related brain activity from resting-state brain dynamics with fmri transformer.Imaging Neuroscience, 3:imag a 00440, 2025

Junbeom Kwon, Jungwoo Seo, Heehwan Wang, Taesup Moon, Shinjae Yoo, and Jiook Cha. Predicting task-related brain activity from resting-state brain dynamics with fmri transformer.Imaging Neuroscience, 3:imag a 00440, 2025. 1, 3

work page 2025
[32]

Zero-shot medical image translation via frequency-guided diffusion models.IEEE transactions on medical imaging, 43(3):980–993, 2023

Yunxiang Li, Hua-Chieh Shao, Xiao Liang, Liyuan Chen, Ruiqi Li, Steve Jiang, Jing Wang, and You Zhang. Zero-shot medical image translation via frequency-guided diffusion models.IEEE transactions on medical imaging, 43(3):980–993, 2023. 1, 3 11 UQ for Dist-to-Dist Flow Matching in Scientific ImagingA PREPRINT

work page 2023
[33]

Diffusion models for out-of-distribution detection in digital pathology.Medical Image Analysis, 93:103088, 2024

Jasper Linmans, Gabriel Raya, Jeroen van der Laak, and Geert Litjens. Diffusion models for out-of-distribution detection in digital pathology.Medical Image Analysis, 93:103088, 2024. 3

work page 2024
[34]

Flow Matching for Generative Modeling

Yaron Lipman, Ricky TQ Chen, Heli Ben-Hamu, Maximilian Nickel, and Matt Le. Flow matching for generative modeling.arXiv preprint arXiv:2210.02747, 2022. 2, 3

work page internal anchor Pith review Pith/arXiv arXiv 2022
[35]

Flow Matching Guide and Code

Yaron Lipman, Marton Havasi, Peter Holderrieth, Neta Shaul, Matt Le, Brian Karrer, Ricky TQ Chen, David Lopez-Paz, Heli Ben-Hamu, and Itai Gat. Flow matching guide and code.arXiv preprint arXiv:2412.06264,

work page internal anchor Pith review Pith/arXiv arXiv
[36]

Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flow

Xingchao Liu, Chengyue Gong, and Qiang Liu. Flow straight and fast: Learning to generate and transfer data with rectified flow.arXiv preprint arXiv:2209.03003, 2022. 4

work page internal anchor Pith review Pith/arXiv arXiv 2022
[37]

Unsupervised out-of-distribution detection with diffusion inpainting

Zhenzhen Liu, Jin Peng Zhou, Yufan Wang, and Kilian Q Weinberger. Unsupervised out-of-distribution detection with diffusion inpainting. InInternational Conference on Machine Learning, pages 22528–22538. PMLR, 2023. 3

work page 2023
[38]

A simple baseline for bayesian uncertainty in deep learning.Advances in neural information processing systems, 32,

Wesley J Maddox, Pavel Izmailov, Timur Garipov, Dmitry P Vetrov, and Andrew Gordon Wilson. A simple baseline for bayesian uncertainty in deep learning.Advances in neural information processing systems, 32,

work page
[39]

SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations

Chenlin Meng, Yutong He, Yang Song, Jiaming Song, Jiajun Wu, Jun-Yan Zhu, and Stefano Ermon. Sdedit: Guided image synthesis and editing with stochastic differential equations.arXiv preprint arXiv:2108.01073,

work page internal anchor Pith review Pith/arXiv arXiv
[40]

Predicting cell morphological responses to per- turbations using generative modeling.Nature Communications, 16(1):505, 2025

Alessandro Palma, Fabian J Theis, and Mohammad Lotfollahi. Predicting cell morphological responses to per- turbations using generative modeling.Nature Communications, 16(1):505, 2025. 8

work page 2025
[41]

Contrastive learning for unpaired image-to- image translation

Taesung Park, Alexei A Efros, Richard Zhang, and Jun-Yan Zhu. Contrastive learning for unpaired image-to- image translation. InEuropean conference on computer vision, pages 319–345. Springer, 2020. 3

work page 2020
[42]

Development of the social brain from age three to twelve years.Nature communications, 9(1):1027, 2018

Hilary Richardson, Grace Lisandrelli, Alexa Riobueno-Naylor, and Rebecca Saxe. Development of the social brain from age three to twelve years.Nature communications, 9(1):1027, 2018. 7, 15

work page 2018
[43]

Diffusion schr¨odinger bridge matching

Yuyang Shi, Valentin De Bortoli, Andrew Campbell, and Arnaud Doucet. Diffusion schr¨odinger bridge matching. Advances in Neural Information Processing Systems, 36:62183–62223, 2023. 3

work page 2023
[44]

Eigenscore: Ood detection using covariance in diffusion models.arXiv preprint arXiv:2510.07206, 2025

Shirin Shoushtari, Yi Wang, Xiao Shi, M Salman Asif, and Ulugbek S Kamilov. Eigenscore: Ood detection using covariance in diffusion models.arXiv preprint arXiv:2510.07206, 2025. 3

work page arXiv 2025
[45]

Score-Based Generative Modeling through Stochastic Differential Equations

Yang Song, Jascha Sohl-Dickstein, Diederik P Kingma, Abhishek Kumar, Stefano Ermon, and Ben Poole. Score- based generative modeling through stochastic differential equations.arXiv preprint arXiv:2011.13456, 2020. 5

work page internal anchor Pith review Pith/arXiv arXiv 2011
[46]

Three forms of stochastic injection for improved distribution-to-distribution generative modeling.arXiv preprint arXiv:2510.06634, 2025

Shiye Su, Yuhui Zhang, Linqi Zhou, Rajesh Ranganath, and Serena Yeung-Levy. Three forms of stochastic injection for improved distribution-to-distribution generative modeling.arXiv preprint arXiv:2510.06634, 2025. 3

work page arXiv 2025
[47]

How to trust your diffusion model: A convex optimization approach to conformal risk control

Jacopo Teneggi, Matthew Tivnan, Web Stayman, and Jeremias Sulam. How to trust your diffusion model: A convex optimization approach to conformal risk control. InInternational Conference on Machine Learning, pages 33940–33960. PMLR, 2023. 3

work page 2023
[48]

Principled probabilistic imaging using diffusion models as plug-and-play priors.Advances in Neural Information Processing Systems, 37:118389–118427, 2024

Zihui Wu, Yu Sun, Yifan Chen, Bingliang Zhang, Yisong Yue, and Katherine Bouman. Principled probabilistic imaging using diffusion models as plug-and-play priors.Advances in Neural Information Processing Systems, 37:118389–118427, 2024. 3

work page 2024
[49]

Measurement-conditioned denoising diffusion probabilistic model for under- sampled medical image reconstruction

Yutong Xie and Quanzheng Li. Measurement-conditioned denoising diffusion probabilistic model for under- sampled medical image reconstruction. InInternational Conference on Medical Image Computing and Computer-Assisted Intervention, pages 655–664. Springer, 2022. 3

work page 2022
[50]

Diffusionad: Norm-guided one-step denoising diffusion for anomaly detection.IEEE Transactions on Pattern Analysis and Machine Intelligence,

Hui Zhang, Zheng Wang, Dan Zeng, Zuxuan Wu, and Yu-Gang Jiang. Diffusionad: Norm-guided one-step denoising diffusion for anomaly detection.IEEE Transactions on Pattern Analysis and Machine Intelligence,

work page
[51]

Understanding failures in out-of-distribution detection with deep generative models

Lily Zhang, Mark Goldstein, and Rajesh Ranganath. Understanding failures in out-of-distribution detection with deep generative models. InInternational Conference on Machine Learning, pages 12427–12436. PMLR, 2021. 8

work page 2021
[52]

Cellflux: Simulating cellular morphology changes via flow matching

Yuhui Zhang, Yuchang Su, Chenyu Wang, Tianhong Li, Zoe Wefers, Jeffrey Nirschl, James Burgess, Daisy Ding, Alejandro Lozano, Emma Lundberg, et al. Cellflux: Simulating cellular morphology changes via flow matching. arXiv preprint arXiv:2502.09775, 2025. 1, 2, 3, 4, 7, 8

work page arXiv 2025
[53]

Guided flows for generative modeling and decision making.arXiv preprint arXiv:2311.13443, 2023

Qinqing Zheng, Matt Le, Neta Shaul, Yaron Lipman, Aditya Grover, and Ricky TQ Chen. Guided flows for generative modeling and decision making.arXiv preprint arXiv:2311.13443, 2023. 4 12 UQ for Dist-to-Dist Flow Matching in Scientific ImagingA PREPRINT

work page arXiv 2023
[54]

Denoising diffusion bridge models.arXiv preprint arXiv:2309.16948, 2023

Linqi Zhou, Aaron Lou, Samar Khanna, and Stefano Ermon. Denoising diffusion bridge models.arXiv preprint arXiv:2309.16948, 2023. 3

work page arXiv 2023
[55]

Unpaired image-to-image translation using cycle- consistent adversarial networks

Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei A Efros. Unpaired image-to-image translation using cycle- consistent adversarial networks. InProceedings of the IEEE international conference on computer vision, pages 2223–2232, 2017. 3 13 UQ for Dist-to-Dist Flow Matching in Scientific ImagingA PREPRINT A Proof of Proposition 4.1 Proof.We prove that t...

work page 2017