Timezone: »

Posterior Meta-Replay for Continual Learning
Christian Henning · Maria Cervera · Francesco D'Angelo · Johannes von Oswald · Regina Traber · Benjamin Ehret · Seijin Kobayashi · Benjamin F. Grewe · João Sacramento

Wed Dec 08 12:30 AM -- 02:00 AM (PST) @

Learning a sequence of tasks without access to i.i.d. observations is a widely studied form of continual learning (CL) that remains challenging. In principle, Bayesian learning directly applies to this setting, since recursive and one-off Bayesian updates yield the same result. In practice, however, recursive updating often leads to poor trade-off solutions across tasks because approximate inference is necessary for most models of interest. Here, we describe an alternative Bayesian approach where task-conditioned parameter distributions are continually inferred from data. We offer a practical deep learning implementation of our framework based on probabilistic task-conditioned hypernetworks, an approach we term posterior meta-replay. Experiments on standard benchmarks show that our probabilistic hypernetworks compress sequences of posterior parameter distributions with virtually no forgetting. We obtain considerable performance gains compared to existing Bayesian CL methods, and identify task inference as our major limiting factor. This limitation has several causes that are independent of the considered sequential setting, opening up new avenues for progress in CL.

Author Information

Christian Henning (ETH Zurich)
Maria Cervera (Swiss Federal Institute of Technology)
Francesco D'Angelo (Swiss Federal Institute of Technology)
Johannes von Oswald (ETH Zurich)
Regina Traber (University of Zurich)
Benjamin Ehret (Swiss Federal Institute of Technology)
Seijin Kobayashi (ETHZ)
Benjamin F. Grewe (ETH Zurich)
João Sacramento (ETH Zurich)

More from the Same Authors