Noradrenergic-inspired gain modulation attenuates the stability gap in joint training
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866917225447817216 |
|---|---|
| author | Rodriguez-Garcia, Alejandro Ghosh, Anindya Ramaswamy, Srikanth |
| author_facet | Rodriguez-Garcia, Alejandro Ghosh, Anindya Ramaswamy, Srikanth |
| contents | Recent work in continual learning has highlighted the stability gap -- a temporary performance drop on previously learned tasks when new ones are introduced. This phenomenon reflects a mismatch between rapid adaptation and strong retention at task boundaries, underscoring the need for optimization mechanisms that balance plasticity and stability over abrupt distribution changes. While optimizers such as momentum-SGD and Adam introduce implicit multi-timescale behavior, they still exhibit pronounced stability gaps. Importantly, these gaps persist even under ideal joint training, making it crucial to study them in this setting to isolate their causes from other sources of forgetting. Motivated by how noradrenergic (neuromodulatory) bursts transiently increase neuronal gain under uncertainty, we introduce a dynamic gain scaling mechanism as a two-timescale optimization technique that balances adaptation and retention by modulating effective learning rates and flattening the local landscape through an effective reparameterization. Across domain- and class-incremental MNIST, CIFAR, and mini-ImageNet benchmarks under task-agnostic joint training, dynamic gain scaling effectively attenuates stability gaps while maintaining competitive accuracy, improving robustness at task transitions. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_14056 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Noradrenergic-inspired gain modulation attenuates the stability gap in joint training Rodriguez-Garcia, Alejandro Ghosh, Anindya Ramaswamy, Srikanth Machine Learning Artificial Intelligence Neurons and Cognition 68T05 Recent work in continual learning has highlighted the stability gap -- a temporary performance drop on previously learned tasks when new ones are introduced. This phenomenon reflects a mismatch between rapid adaptation and strong retention at task boundaries, underscoring the need for optimization mechanisms that balance plasticity and stability over abrupt distribution changes. While optimizers such as momentum-SGD and Adam introduce implicit multi-timescale behavior, they still exhibit pronounced stability gaps. Importantly, these gaps persist even under ideal joint training, making it crucial to study them in this setting to isolate their causes from other sources of forgetting. Motivated by how noradrenergic (neuromodulatory) bursts transiently increase neuronal gain under uncertainty, we introduce a dynamic gain scaling mechanism as a two-timescale optimization technique that balances adaptation and retention by modulating effective learning rates and flattening the local landscape through an effective reparameterization. Across domain- and class-incremental MNIST, CIFAR, and mini-ImageNet benchmarks under task-agnostic joint training, dynamic gain scaling effectively attenuates stability gaps while maintaining competitive accuracy, improving robustness at task transitions. |
| title | Noradrenergic-inspired gain modulation attenuates the stability gap in joint training |
| topic | Machine Learning Artificial Intelligence Neurons and Cognition 68T05 |
| url | https://arxiv.org/abs/2507.14056 |