Papers › Towards Redundancy-Free Sub-networks in Continual Learning

Towards Redundancy-Free Sub-networks in Continual Learning

1 Dec 2023arXiv:2312.00840archive 2025-07-28

Cheng Chen, Jingkuan Song, Lianli Gao, Heng Tao Shen

Catastrophic Forgetting (CF) is a prominent issue in continual learning. Parameter isolation addresses this challenge by masking a sub-network for each task to mitigate interference with old tasks. However, these sub-networks are constructed relying on weight magnitude, which does not necessarily correspond to the importance of weights, resulting in maintaining unimportant weights and constructing redundant sub-networks. To overcome this limitation, inspired by information bottleneck, which removes redundancy between adjacent network layers, we propose \textbf{\underline{I}nformation \underline{B}ottleneck \underline{M}asked sub-network (IBM)} to eliminate redundancy within sub-networks. Specifically, IBM accumulates valuable information into essential weights to construct redundancy-free sub-networks, not only effectively mitigating CF by freezing the sub-networks but also facilitating new tasks training through the transfer of valuable knowledge. Additionally, IBM decomposes hidden representations to automate the construction process and make it flexible. Extensive experiments demonstrate that IBM consistently outperforms state-of-the-art methods. Notably, IBM surpasses the state-of-the-art parameter isolation method with a 70\% reduction in the number of parameters within sub-networks and an 80\% decrease in training time.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

zackschen/IBM-Net officialmentioned in paperpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Continual Learning

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Continual Learning CIFAR-100 AlexNet - 300 Epoch IBM Accuracy 82.69 #1 of 1 Archive leaderboard report
Continual Learning CIFAR-100 ResNet-18 - 300 Epochs IBM Accuracy 88.15 #1 of 1 Archive leaderboard report
Continual Learning MiniImageNet ResNet-18 - 300 Epochs IBM Accuracy 53.9 #1 of 1 Archive leaderboard report
Continual Learning TinyImageNet ResNet-18 - 300 Epochs IBM Accuracy 52.38 #1 of 1 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections