Rapid learning with phase-change memory-based in-memory computing through learning-to-learn
Abstract There is a growing demand for low-power, autonomously learning artificial intelligence (AI) systems that can be applied at the edge and rapidly adapt to the specific situation at deployment site. However, current AI models struggle in such scenarios, often requiring extensive fine-tuning, c...
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-02-01
|
| Series: | Nature Communications |
| Online Access: | https://doi.org/10.1038/s41467-025-56345-4 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832571504890478592 |
|---|---|
| author | Thomas Ortner Horst Petschenig Athanasios Vasilopoulos Roland Renner Špela Brglez Thomas Limbacher Enrique Piñero Alejandro Linares-Barranco Angeliki Pantazi Robert Legenstein |
| author_facet | Thomas Ortner Horst Petschenig Athanasios Vasilopoulos Roland Renner Špela Brglez Thomas Limbacher Enrique Piñero Alejandro Linares-Barranco Angeliki Pantazi Robert Legenstein |
| author_sort | Thomas Ortner |
| collection | DOAJ |
| description | Abstract There is a growing demand for low-power, autonomously learning artificial intelligence (AI) systems that can be applied at the edge and rapidly adapt to the specific situation at deployment site. However, current AI models struggle in such scenarios, often requiring extensive fine-tuning, computational resources, and data. In contrast, humans can effortlessly adjust to new tasks by transferring knowledge from related ones. The concept of learning-to-learn (L2L) mimics this process and enables AI models to rapidly adapt with only little computational effort and data. In-memory computing neuromorphic hardware (NMHW) is inspired by the brain’s operating principles and mimics its physical co-location of memory and compute. In this work, we pair L2L with in-memory computing NMHW based on phase-change memory devices to build efficient AI models that can rapidly adapt to new tasks. We demonstrate the versatility of our approach in two scenarios: a convolutional neural network performing image classification and a biologically-inspired spiking neural network generating motor commands for a real robotic arm. Both models rapidly learn with few parameter updates. Deployed on the NMHW, they perform on-par with their software equivalents. Moreover, meta-training of these models can be performed in software with high-precision, alleviating the need for accurate hardware models. |
| format | Article |
| id | doaj-art-38513e14accc4f1db96d382521d637ca |
| institution | Kabale University |
| issn | 2041-1723 |
| language | English |
| publishDate | 2025-02-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Nature Communications |
| spelling | doaj-art-38513e14accc4f1db96d382521d637ca2025-02-02T12:31:21ZengNature PortfolioNature Communications2041-17232025-02-0116111610.1038/s41467-025-56345-4Rapid learning with phase-change memory-based in-memory computing through learning-to-learnThomas Ortner0Horst Petschenig1Athanasios Vasilopoulos2Roland Renner3Špela Brglez4Thomas Limbacher5Enrique Piñero6Alejandro Linares-Barranco7Angeliki Pantazi8Robert Legenstein9IBM Research Europe - ZurichInstitute of Machine Learning and Neural Computation, Graz University of TechnologyIBM Research Europe - ZurichInstitute of Machine Learning and Neural Computation, Graz University of TechnologyInstitute of Machine Learning and Neural Computation, Graz University of TechnologyInstitute of Machine Learning and Neural Computation, Graz University of TechnologyRobotics and Technology of Computers, SCORE Laboratory EPS-ETSII, Universidad de SevillaRobotics and Technology of Computers, SCORE Laboratory EPS-ETSII, Universidad de SevillaIBM Research Europe - ZurichInstitute of Machine Learning and Neural Computation, Graz University of TechnologyAbstract There is a growing demand for low-power, autonomously learning artificial intelligence (AI) systems that can be applied at the edge and rapidly adapt to the specific situation at deployment site. However, current AI models struggle in such scenarios, often requiring extensive fine-tuning, computational resources, and data. In contrast, humans can effortlessly adjust to new tasks by transferring knowledge from related ones. The concept of learning-to-learn (L2L) mimics this process and enables AI models to rapidly adapt with only little computational effort and data. In-memory computing neuromorphic hardware (NMHW) is inspired by the brain’s operating principles and mimics its physical co-location of memory and compute. In this work, we pair L2L with in-memory computing NMHW based on phase-change memory devices to build efficient AI models that can rapidly adapt to new tasks. We demonstrate the versatility of our approach in two scenarios: a convolutional neural network performing image classification and a biologically-inspired spiking neural network generating motor commands for a real robotic arm. Both models rapidly learn with few parameter updates. Deployed on the NMHW, they perform on-par with their software equivalents. Moreover, meta-training of these models can be performed in software with high-precision, alleviating the need for accurate hardware models.https://doi.org/10.1038/s41467-025-56345-4 |
| spellingShingle | Thomas Ortner Horst Petschenig Athanasios Vasilopoulos Roland Renner Špela Brglez Thomas Limbacher Enrique Piñero Alejandro Linares-Barranco Angeliki Pantazi Robert Legenstein Rapid learning with phase-change memory-based in-memory computing through learning-to-learn Nature Communications |
| title | Rapid learning with phase-change memory-based in-memory computing through learning-to-learn |
| title_full | Rapid learning with phase-change memory-based in-memory computing through learning-to-learn |
| title_fullStr | Rapid learning with phase-change memory-based in-memory computing through learning-to-learn |
| title_full_unstemmed | Rapid learning with phase-change memory-based in-memory computing through learning-to-learn |
| title_short | Rapid learning with phase-change memory-based in-memory computing through learning-to-learn |
| title_sort | rapid learning with phase change memory based in memory computing through learning to learn |
| url | https://doi.org/10.1038/s41467-025-56345-4 |
| work_keys_str_mv | AT thomasortner rapidlearningwithphasechangememorybasedinmemorycomputingthroughlearningtolearn AT horstpetschenig rapidlearningwithphasechangememorybasedinmemorycomputingthroughlearningtolearn AT athanasiosvasilopoulos rapidlearningwithphasechangememorybasedinmemorycomputingthroughlearningtolearn AT rolandrenner rapidlearningwithphasechangememorybasedinmemorycomputingthroughlearningtolearn AT spelabrglez rapidlearningwithphasechangememorybasedinmemorycomputingthroughlearningtolearn AT thomaslimbacher rapidlearningwithphasechangememorybasedinmemorycomputingthroughlearningtolearn AT enriquepinero rapidlearningwithphasechangememorybasedinmemorycomputingthroughlearningtolearn AT alejandrolinaresbarranco rapidlearningwithphasechangememorybasedinmemorycomputingthroughlearningtolearn AT angelikipantazi rapidlearningwithphasechangememorybasedinmemorycomputingthroughlearningtolearn AT robertlegenstein rapidlearningwithphasechangememorybasedinmemorycomputingthroughlearningtolearn |