mlsanityData to 7 Oct 2026

CellFluxRL: Biologically-Constrained Virtual Cell Modeling via Reinforcement Learning

arXiv 2603.21743 ↗First posted 23 Mar 2026cs.LG q-bio.QMcited by 1

Not on the map. Why?

Abstract (arXiv)

Building virtual cells with generative models to simulate cellular behavior in silico is emerging as a promising paradigm for accelerating drug discovery. However, prior image-based generative approaches can produce implausible cell images that violate basic physical and biological constraints. To address this, we propose to post-train virtual cell models with reinforcement learning (RL), leveraging biologically meaningful evaluators as reward functions. We design seven rewards spanning three categories-biological function, structural validity, and morphological correctness-and optimize the state-of-the-art CellFlux model to yield CellFluxRL. CellFluxRL consistently improves over CellFlux across all rewards, with further performance boosts from test-time scaling. Overall, our results present a virtual cell modeling framework that enforces physically-based constraints through RL, advancing beyond "visually realistic" generations towards "biologically meaningful" ones.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 34 of its 53 references that are in the map. First posted on arXiv 23 Mar 2026.

Builds on

34 of 53 references.

References · 34

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  2. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  3. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  4. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,764
  5. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,005
  6. 2022Flow Matching for Generative Modelingcited by 5,426
  7. 2019Generative Modeling by Estimating Gradients of the Data Distributioncited by 4,067
  8. 2023Let's Verify Step by Stepcited by 3,598
  9. 2022Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flowcited by 3,550
  10. 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameterscited by 1,928
  11. 2023Training Diffusion Models with Reinforcement Learningcited by 939
  12. 2025Flow-GRPO: Training Flow Matching Models via Online RLcited by 605
22 more
  1. 2023DPOK: Reinforcement Learning for Fine-tuning Text-to-Image Diffusion Modelscited by 431
  2. 2025DanceGRPO: Unleashing GRPO on Visual Generationcited by 388
  3. 2022DayDreamer: World Models for Physical Robot Learningcited by 372
  4. 2024Flow Matching Guide and Codecited by 291
  5. 2025Improving Video Generation with Human Feedbackcited by 267
  6. 2023Diffusion-based Generation, Optimization, and Planning in 3D Scenescited by 237
  7. 2025Inference-Time Scaling for Diffusion Models beyond Scaling Denoising Stepscited by 225
  8. 2024How Far is Video Generation from World Model: A Physical Law Perspectivecited by 204
  9. 2025DiffusionNFT: Online Diffusion Reinforcement with Forward Processcited by 204
  10. 2018Cell Detection with Star-convex Polygonscited by 182
  11. 2025MixGRPO: Unlocking Flow-based GRPO Efficiency with Mixed ODE-SDEcited by 168
  12. 2024Towards World Simulator: Crafting Physical Commonsense-Based Benchmark for Video Generationcited by 167
  13. 2022TransDreamer: Reinforcement Learning with Transformer World Modelscited by 143
  14. 2023VideoDirectorGPT: Consistent Multi-scene Video Generation via LLM-Guided Planningcited by 138
  15. 2024PhyScene: Physically Interactable 3D Scene Synthesis for Embodied AIcited by 116
  16. 2024How to Build the Virtual Cell with Artificial Intelligence: Priorities and Opportunitiescited by 50
  17. 2023Modelling Cellular Perturbations with the Sparse Additive Mechanism Shift Variational Autoencodercited by 35
  18. 2025PhyGDPO: Physics-Aware Groupwise Direct Preference Optimization for Physically Consistent Text-to-Video Generationcited by 22
  19. 2025CellFlux: Simulating Cellular Morphology Changes via Flow Matchingcited by 19
  20. 2024Compositional 3D-aware Video Generation with LLM Directorcited by 18
  21. 2023PhenDiff: Revealing Subtle Phenotypes with Diffusion Models in Real Imagescited by 10
  22. 2025RDPO: Real Data Preference Optimization for Physics Consistency Video Generationcited by 8

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026PerturbCellRL: Aligning Distributions and Grounding Biology via Post-Training Perturbation Generatorscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Feedback Efficient Online Fine-Tuning of Diffusion Models
  2. 2025VCWorld: A Biological World Model for Virtual Cell Simulation
  3. 2025CellForge: Agentic Design of Virtual Cell Models
  4. 2026VCLMU: Mechanism-Centric Virtual Cell World Modeling for Perturbation Response
  5. 2024Bridging Model-Based Optimization and Generative Modeling via Conservative Fine-Tuning of Diffusion Models
  6. 2025Cell-o1: Training LLMs to Solve Single-Cell Reasoning Puzzles with Reinforcement Learning
  7. 2025MorphGen: Controllable and Morphologically Plausible Generative Cell-Imaging
  8. 2025Iterative Distillation for Reward-Guided Fine-Tuning of Diffusion Models in Biomolecular Design