Giuseppe Serra
@serpep.bsky.social
Machine Learning Researcher | Trying to unbox black-boxes.
Reposted by Giuseppe Serra
We got accepted at TMLR!
What characteristics of the samples consistently alleviate catastrophic forgetting in memory-based online continual learning?
@serpep.bsky.social, Ben Werner, and Florian Buettner investigated this question under an uncertainty lens.
📜: openreview.net/forum?id=dcz...
What characteristics of the samples consistently alleviate catastrophic forgetting in memory-based online continual learning?
@serpep.bsky.social, Ben Werner, and Florian Buettner investigated this question under an uncertainty lens.
📜: openreview.net/forum?id=dcz...
February 20, 2025 at 12:27 PM
We got accepted at TMLR!
What characteristics of the samples consistently alleviate catastrophic forgetting in memory-based online continual learning?
@serpep.bsky.social, Ben Werner, and Florian Buettner investigated this question under an uncertainty lens.
📜: openreview.net/forum?id=dcz...
What characteristics of the samples consistently alleviate catastrophic forgetting in memory-based online continual learning?
@serpep.bsky.social, Ben Werner, and Florian Buettner investigated this question under an uncertainty lens.
📜: openreview.net/forum?id=dcz...
Let's inaugurate this new account with great news! In our #ICLR2025 paper, we tackle the *online* problem in federated continual learning. Via an uncertainty-aware memory-based approach, we consider settings where retaining the full dataset locally is not possible.
We got accepted at #ICLR2025!
@serpep.bsky.social and Florian Buettner formalized a novel scenario for *online* federated continual learning and introduced an effective memory-based baseline that combines uncertainty-aware updates with random replay.
📄Read more: openreview.net/forum?id=f65...
@serpep.bsky.social and Florian Buettner formalized a novel scenario for *online* federated continual learning and introduced an effective memory-based baseline that combines uncertainty-aware updates with random replay.
📄Read more: openreview.net/forum?id=f65...
February 17, 2025 at 7:33 AM
Let's inaugurate this new account with great news! In our #ICLR2025 paper, we tackle the *online* problem in federated continual learning. Via an uncertainty-aware memory-based approach, we consider settings where retaining the full dataset locally is not possible.