![[2602.22884] Unsupervised Continual Learning for Amortized Bayesian Inference](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.22884] Unsupervised Continual Learning for Amortized Bayesian Inference
Summary
This article presents a novel framework for Unsupervised Continual Learning in Amortized Bayesian Inference, addressing performance issues under model misspecification and catastrophic forgetting.
Why It Matters
The research is significant as it proposes solutions for enhancing the robustness of Amortized Bayesian Inference models in dynamic environments, which is crucial for real-world applications where data distributions can shift over time. This work could lead to more reliable machine learning systems in various fields.
Key Takeaways
- Introduces a continual learning framework to improve Amortized Bayesian Inference.
- Proposes two adaptation strategies to mitigate catastrophic forgetting.
- Demonstrates significant performance improvements over standard training methods.
Statistics > Machine Learning arXiv:2602.22884 (stat) [Submitted on 26 Feb 2026] Title:Unsupervised Continual Learning for Amortized Bayesian Inference Authors:Aayush Mishra, Šimon Kucharský, Paul-Christian Bürkner View a PDF of the paper titled Unsupervised Continual Learning for Amortized Bayesian Inference, by Aayush Mishra and 2 other authors View PDF HTML (experimental) Abstract:Amortized Bayesian Inference (ABI) enables efficient posterior estimation using generative neural networks trained on simulated data, but often suffers from performance degradation under model misspecification. While self-consistency (SC) training on unlabeled empirical data can enhance network robustness, current approaches are limited to static, single-task settings and fail to handle sequentially arriving data or distribution shifts. We propose a continual learning framework for ABI that decouples simulation-based pre-training from unsupervised sequential SC fine-tuning on real-world data. To address the challenge of catastrophic forgetting, we introduce two adaptation strategies: (1) SC with episodic replay, utilizing a memory buffer of past observations, and (2) SC with elastic weight consolidation, which regularizes updates to preserve task-critical parameters. Across three diverse case studies, our methods significantly mitigate forgetting and yield posterior estimates that outperform standard simulation-based training, achieving estimates closer to MCMC reference, providing a viable pa...
