skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: Reducing Catastrophic Forgetting With Associative Learning: A Lesson From Fruit Flies
Abstract Catastrophic forgetting remains an outstanding challenge in continual learning. Recently, methods inspired by the brain, such as continual representation learning and memory replay, have been used to combat catastrophic forgetting. Associative learning (retaining associations between inputs and outputs, even after good representations are learned) plays an important function in the brain; however, its role in continual learning has not been carefully studied. Here, we identified a two-layer neural circuit in the fruit fly olfactory system that performs continual associative learning between odors and their associated valences. In the first layer, inputs (odors) are encoded using sparse, high-dimensional representations, which reduces memory interference by activating nonoverlapping populations of neurons for different odors. In the second layer, only the synapses between odor-activated neurons and the odor’s associated output neuron are modified during learning; the rest of the weights are frozen to prevent unrelated memories from being overwritten. We prove theoretically that these two perceptron-like layers help reduce catastrophic forgetting compared to the original perceptron algorithm, under continual learning. We then show empirically on benchmark data sets that this simple and lightweight architecture outperforms other popular neural-inspired algorithms when also using a two-layer feedforward architecture. Overall, fruit flies evolved an efficient continual associative learning algorithm, and circuit mechanisms from neuroscience can be translated to improve machine computation.  more » « less
Award ID(s):
2211386
PAR ID:
10517683
Author(s) / Creator(s):
; ;
Publisher / Repository:
MIT Press
Date Published:
Journal Name:
Neural Computation
Volume:
35
Issue:
11
ISSN:
0899-7667
Page Range / eLocation ID:
1797 to 1819
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; ; ; (, European Conference on Computer Vision)
    null (Ed.)
    People learn throughout life. However, incrementally updating conventional neural networks leads to catastrophic forgetting. A common remedy is replay, which is inspired by how the brain consolidates memory. Replay involves fine-tuning a network on a mixture of new and old instances. While there is neuroscientific evidence that the brain replays compressed memories, existing methods for convolutional networks replay raw images. Here, we propose REMIND, a brain-inspired approach that enables efficient replay with compressed representations. REMIND is trained in an online manner, meaning it learns one example at a time, which is closer to how humans learn. Under the same constraints, REMIND outperforms other methods for incremental class learning on the ImageNet ILSVRC-2012 dataset. We probe REMIND’s robustness to data ordering schemes known to induce catastrophic forgetting. We demonstrate REMIND’s generality by pioneering online learning for Visual Question Answering (VQA). 
    more » « less
  2. ; ; ; (, Nature Communications)
    Abstract Artificial neural networks are known to suffer from catastrophic forgetting: when learning multiple tasks sequentially, they perform well on the most recent task at the expense of previously learned tasks. In the brain, sleep is known to play an important role in incremental learning by replaying recent and old conflicting memory traces. Here we tested the hypothesis that implementing a sleep-like phase in artificial neural networks can protect old memories during new training and alleviate catastrophic forgetting. Sleep was implemented as off-line training with local unsupervised Hebbian plasticity rules and noisy input. In an incremental learning framework, sleep was able to recover old tasks that were otherwise forgotten. Previously learned memories were replayed spontaneously during sleep, forming unique representations for each class of inputs. Representational sparseness and neuronal activity corresponding to the old tasks increased while new task related activity decreased. The study suggests that spontaneous replay simulating sleep-like dynamics can alleviate catastrophic forgetting in artificial neural networks. 
    more » « less
  3. ; ; ; ; (, IEEE)
    Inspired by the success of Self-Supervised Learning (SSL) in learning visual representations from unlabeled data, a few recent works have studied SSL in the context of Continual Learning (CL), where multiple tasks are learned sequentially, giving rise to a new paradigm, namely Self-Supervised Continual Learning (SSCL). It has been shown that the SSCL outperforms Supervised Continual Learning (SCL) as the learned representations are more informative and robust to catastrophic forgetting. However, building upon the training process of SSL, prior SSCL studies involve training all the parameters for each task, resulting to prohibitively high training cost. In this work, we first analyze the training time and memory consumption and reveals that the backward gradient calculation is the bottleneck. Moreover, by investigating the task correlations in SSCL, we further discover an interesting phenomenon that, with the SSL-learned background model, the intermediate features are highly correlated between tasks. Based on these new finding, we propose a new SSCL method with layer-wise freezing which progressively freezes partial layers with the highest correlation ratios for each task to improve training computation efficiency and memory efficiency. Extensive experiments across multiple datasets are performed, where our proposed method shows superior performance against the SoTA SSCL methods under various SSL frameworks. For example, compared to LUMP, our method achieves 1.18x, 1.15x, and 1.2x GPU training time reduction, 1.65x, 1.61x, and 1.6x memory reduction, 1.46x, 1.44x, and 1.46x backward FLOPs reduction, and 1.31%/1.98%/1.21% forgetting reduction without accuracy degradation on three datasets, respectively. 
    more » « less
  4. ; ; ; ; ; ; ; ; ; ; et al (, The Journal of Neuroscience)
    Larvae of the fruit flyDrosophila melanogasterare a powerful study case for understanding the neural circuits underlying behavior. Indeed, the numerical simplicity of the larval brain has permitted the reconstruction of its synaptic connectome, and genetic tools for manipulating single, identified neurons allow neural circuit function to be investigated with relative ease and precision. We focus on one of the most complex neurons in the brain of the larva (of either sex), the GABAergic anterior paired lateral neuron (APL). Using behavioral and connectomic analyses, optogenetics, Ca2+imaging, and pharmacology, we study how APL affects associative olfactory memory. We first provide a detailed account of the structure, regional polarity, connectivity, and metamorphic development of APL, and further confirm that optogenetic activation of APL has an inhibiting effect on its main targets, the mushroom body Kenyon cells. All these findings are consistent with the previously identified function of APL in the sparsening of sensory representations. To our surprise, however, we found that optogenetically activating APL can also have a strong rewarding effect. Specifically, APL activation together with odor presentation establishes an odor-specific, appetitive, associative short-term memory, whereas naive olfactory behavior remains unaffected. An acute, systemic inhibition of dopamine synthesis as well as an ablation of the dopaminergic pPAM neurons impair reward learning through APL activation. Our findings provide a study case of complex circuit function in a numerically simple brain, and suggest a previously unrecognized capacity of central-brain GABAergic neurons to engage in dopaminergic reinforcement. SIGNIFICANCE STATEMENTThe single, identified giant anterior paired lateral (APL) neuron is one of the most complex neurons in the insect brain. It is GABAergic and contributes to the sparsening of neuronal activity in the mushroom body, the memory center of insects. We provide the most detailed account yet of the structure of APL in larvalDrosophilaas a neurogenetically accessible study case. We further reveal that, contrary to expectations, the experimental activation of APL can exert a rewarding effect, likely via dopaminergic reward pathways. The present study both provides an example of unexpected circuit complexity in a numerically simple brain, and reports an unexpected effect of activity in central-brain GABAergic circuits. 
    more » « less
  5. ; ; ; (, CVPR Workshop on Continual Learning in Computer Vision (CLVISION))
    Supervised Continual learning involves updating a deep neural network (DNN) from an ever-growing stream of labeled data. While most work has focused on overcoming catastrophic forgetting, one of the major motivations behind continual learning is being able to efficiently update a network with new information, rather than retraining from scratch on the training dataset as it grows over time. Despite recent continual learning methods largely solving the catastrophic forgetting problem, there has been little attention paid to the efficiency of these algorithms. Here, we study recent methods for incremental class learning and illustrate that many are highly inefficient in terms of compute, memory, and storage. Some methods even require more compute than training from scratch! We argue that for continual learning to have real-world applicability, the research community cannot ignore the resources used by these algorithms. There is more to continual learning than mitigating catastrophic forgetting. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email