Tokenised Flow Matching for Hierarchical Simulation Based Inference

1. 1Apply likelihood factorisation to hierarchical models to reduce simulation costs
2. 2Use tokenised flow matching to enable efficient simulation-based inference
3. 3Implement hierarchical simulation-based inference with shared global parameters and exchangeable site-level parameters
4. 4Evaluate the performance of tokenised flow matching using metrics such as simulation efficiency and inference accuracy
5. 5Compare the results with existing hierarchical SBI approaches to assess the benefits of the new method

Data scientists and machine learning engineers working on simulation-based inference projects can benefit from this approach to improve efficiency and reduce computational costs

💡 Tokenised flow matching can be used to exploit the structure of hierarchical models and improve simulation efficiency in simulation-based inference

💡 Improve simulation efficiency in hierarchical settings with tokenised flow matching for simulation-based inference!