Memory-Integrated Reconfigurable Adapters: A Unified Framework for Settings with Multiple Tasks
Susmit Agrawal, Krishn Vishwas Kher, Saksham Mittal, Swarnim Maheshwari, Vineeth N. Balasubramanian

TL;DR
This paper introduces MIRA, a unified neural architecture combining associative memories with adapters, enabling rapid multi-task adaptation and knowledge retention, inspired by biological neural mechanisms.
Contribution
MIRA is the first framework to integrate associative memory modules with adapters for unified domain generalization and continual learning.
Findings
Achieves state-of-the-art out-of-distribution accuracy in domain generalization.
Outperforms existing continual learning architectures in incremental tasks.
Enhances adaptability and knowledge retention in multi-task settings.
Abstract
Organisms constantly pivot between tasks such as evading predators, foraging, traversing rugged terrain, and socializing, often within milliseconds. Remarkably, they preserve knowledge of once-learned environments sans catastrophic forgetting, a phenomenon neuroscientists hypothesize, is due to a singular neural circuitry dynamically overlayed by neuromodulatory agents such as dopamine and acetylcholine. In parallel, deep learning research addresses analogous challenges via domain generalization (DG) and continual learning (CL), yet these methods remain siloed, despite the brains ability to perform them seamlessly. In particular, prior work has not explored architectures involving associative memories (AMs), which are an integral part of biological systems, to jointly address these tasks. We propose Memory-Integrated Reconfigurable Adapters (MIRA), a unified framework that integrates…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
Taxonomy
TopicsDomain Adaptation and Few-Shot Learning · Ferroelectric and Negative Capacitance Devices · Advanced Memory and Neural Computing
