Probing Representation Forgetting in Supervised and Unsupervised Continual Learning
Open accessThe DOI matches its Crossref record; 15 of the 15 references it cites were found in a registry or an index.
Abstract
Continual Learning (CL) research typically focuses on tackling the phenomenon of catastrophic forgetting in neural networks. Catastrophic forgetting is associated with an abrupt loss of knowledge previously learned by a model when the task, or more broadly the data distribution, being trained on changes. In supervised learning problems this forgetting, resulting from a change in the model's representation, is typically measured or observed by evaluating the decrease in old task performance. However, a model's representation can change without losing knowledge about prior tasks. In this work we consider the concept of representation forgetting, observed by using the difference in performance of an optimal linear classifier before and after a new task is introduced. Using this tool we revisit a number of standard continual learning benchmarks and observe that, through this lens, model representations trained without any explicit control for forgetting often experience small representation forgetting and can sometimes be comparable to methods which explicitly control for forgetting, especially in longer task sequences. We also show that representation forgetting can lead to new insights on the effect of model capacity and loss function used in continual learning. Based on our results, we show that a simple yet competitive approach is to learn representations continually with standard supervised contrastive learning while constructing prototypes of class samples when queried on old samples.11The code to reproduce our results is publicly available at: https://github.com/rezazzr/Probing-Representation-Forgetting
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474
Overcoming catastrophic forgetting in neural networks
Kirkpatrick, J., Pascanu, R., Rabinowitz, N., Veness, J., Desjardins, G., Rusu, A. A., Milan, K., Quan, J., Ramalho, T., Grabska-Barwinska, A., et al. (2017) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1611835114 · Cited by 7,899
Li, Z., & Hoiem, D. (2018) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2017.2773081 · Cited by 4,202
iCaRL: Incremental Classifier and Representation Learning
Rebuffi, S.-A., Kolesnikov, A., Sperl, G., & Lampert, C. H. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.587 · Cited by 3,988
Catastrophic Interference in Connectionist Networks: The Sequential Learning Problem
McCloskey, M., & Cohen, N. J. (1989) · Psychology of Learning and Motivation · doi:10.1016/s0079-7421(08)60536-8 · Cited by 3,770
Supervised Contrastive Learning
Khosla, P., Teterwak, P., Wang, C., Sarna, A., Tian, Y., Isola, P., Maschinot, A., Liu, C., & Krishnan, D. (2020) · arXiv · doi:10.48550/arxiv.2004.11362 · Cited by 3,569 · cited 5×
“As well they are becoming a popular alternative to cross-entropy (CE) in the supervised setting [12, 21], referred to as SupCon.”
BackgroundMethodInfluentialAutomated Flower Classification over a Large Number of Classes
Nilsback, M.-E., & Zisserman, A. (2008) · 2008 Sixth Indian Conference on Computer Vision, Graphics & Image Processing · doi:10.1109/icvgip.2008.47 · Cited by 3,442 · cited 2×
“…samples over 67 classes), Caltech-UCSD Birds (CUB) [47] for classification of bird species (6,033 samples over 200 classes), and Oxford Flowers [35] for flower classification (2,040 samples over 102 Observed and LP Accuracy SplitMiniImageNet We compare observed accuracy and Linear probe accuracy…”
MethodTutorial on maximum likelihood estimation
Myung, I. J. (2003) · Journal of Mathematical Psychology · doi:10.1016/s0022-2496(02)00028-7 · Cited by 1,802
Memory Aware Synapses: Learning What (not) to Forget
Aljundi, R., Babiloni, F., Elhoseiny, M., Rohrbach, M., & Tuytelaars, T. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-01219-9_9 · Cited by 1,590
Self-citationGradient Episodic Memory for Continual Learning
Lopez-Paz, D., & Ranzato, M. (2017) · arXiv · doi:10.48550/arxiv.1706.08840 · Cited by 1,403
Cited by
A Comprehensive Survey of Continual Learning: Theory, Method and Application
Wang, L., Zhang, X., Su, H., & Zhu, J. (2024) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2024.3367329 · Cited by 1,113
“In particular, the benefits for downstream continual learning tend to be more apparent when pre-training with larger data size [316], [352], larger model size [352] and contrastive loss [83], [125].”
BackgroundAn Empirical Study of Catastrophic Forgetting in Large Language Models During Continual Fine-Tuning
Luo, Y., Yang, Z., Meng, F., Li, Y., Zhou, J., & Zhang, Y. (2025) · IEEE Transactions on Audio, Speech and Language Processing · doi:10.1109/taslpro.2025.3606231 · Cited by 136
“However, Davari et al. (2022) [22] discovered that even when the model performance on previously learned tasks is preserved, the representations still suffer from significant drift due to parameter updates.”
An Empirical Study of Catastrophic Forgetting in Large Language Models During Continual Fine-tuning
Luo, Y., Yang, Z., Meng, F., Li, Y., Zhou, J., & Zhang, Y. (2023) · arXiv · doi:10.48550/arxiv.2308.08747 · Cited by 135
Catastrophic Forgetting in Deep Learning: A Comprehensive Taxonomy
Aleixo, E. L., Colonna, J. G., Cristo, M., & Fernandes, E. (2023) · arXiv · doi:10.48550/arxiv.2312.10549 · Cited by 54
“Distance-based approaches benefit from using cosine distance classifiers, which have shown greater resilience to CF compared to softmax classifiers [Simon et al ., 2021; Davari et al ., 2022].”
MethodCatastrophic Forgetting in Deep Learning: A Comprehensive Taxonomy
Aleixo, E. L., Colonna, J. G., Cristo, M., & Fernandes, E. (2024) · Journal of the Brazilian Computer Society · doi:10.5753/jbcs.2024.3966 · Cited by 54
Metadata
- Authors
- Davari, M., Asadi, N., Mudur, S., Aljundi, R., & Belilovsky, E.
- Venue
- 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
- Publisher
- IEEE
- Published
- 2022-06
- Pages
- 16691-16700
- ISSN
- 1063-6919 (ISSN-L)
- OpenAlex
- W4312553254
- Cited by
- 55
- References
- 15
- FWCI
- 7.04
- Type
- conference paper
- Language
- en
- Open access
- green
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.