Stefan-CL reframes continual learning as a multiphase moving-boundary problem, using a latent-heat parameter to expand a protected solid region of learned knowledge and achieve near-zero forgetting without storing raw data.
A coupled kolmogorov--arnold network and level-set framework for evolving interfaces
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Continual Learning as a Multiphase Moving-Boundary Problem
Stefan-CL reframes continual learning as a multiphase moving-boundary problem, using a latent-heat parameter to expand a protected solid region of learned knowledge and achieve near-zero forgetting without storing raw data.