Uncovering and Fixing Collider Bias in Bayesian PINNs
Quick summary
arXiv:2610.11737v1 Announce Type: cross Abstract: Bayesian physics-informed neural networks (B-PINNs) are a popular framework for parameter and state inference from sparse or noisy observations. They are commonly formulated via a collider structure, in which physical and trajectory parameters are assumed to be a priori independent and become coupled through virtual likelihoods on differential-equation residuals that enforce physical consistency. We show that this modeling choice can induce severe systematic bias in the posterior over physical parameters: even when the prior is favorably center
Key takeaways
- arXiv:2610.11737v1 Announce Type: cross Abstract: Bayesian physics-informed neural networks (B-PINNs) are a popular framework for parameter and state inference from sparse or noisy observations.
- They are commonly formulated via a collider structure, in which physical and trajectory parameters are assumed to be a priori independent and become coupled through virtual likelihoods on differential-equation residuals that enforce physical consistency.
- We show that this modeling choice can induce severe systematic bias in the posterior over physical parameters: even when the prior is favorably center
Why it matters
The importance of “Uncovering and Fixing Collider Bias in Bayesian PINNs” will be measured by what changes in practice. User behavior, access conditions, verifiable performance and responsible-use outcomes are the signals worth following.
Member comments