Two-shot learning of multiple strange attractors

Daniel Köglmayr, Miralem Spahic, Andrew Flynn, Christoph Räth · 2026

Read Two-shot learning of multiple strange attractors by Daniel Köglmayr, Miralem Spahic, Andrew Flynn, Christoph Räth (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-21117.txt.

Description

Daniel Köglmayr and three co-authors teach a machine to remember several chaotic patterns after just two glimpses. Their 2026 paper studies two-shot learning of multiple strange attractors. A strange attractor is the tangled shape a chaotic system traces over time. The brain combines short- and long-term memory to store many pieces of information. The authors take inspiration from that design. They extend a recent machine learning technique that does the same.

Their system pairs a next-generation reservoir computer with extremely randomized trees. The reservoir acts as short-term memory; the trees hold the long-term store. Together they process, store, and recall multiple attractors. Learning each from only two examples is the hard part. Success hints at machines that memorize dynamics like brains do. The work joins neuroscience ideas with chaotic time-series learning.

Catalog record

CREATORDaniel Köglmayr, Miralem Spahic, Andrew Flynn, Christoph Räth
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE75 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-21117

Browse the full library · About the Tower of Babel