ReTune proves that restarting a contractive truncated unrolled solver and backpropagating only through the final block yields gradients arbitrarily close to the deep-equilibrium hypergradient.
NTIRE 2017 challenge on single image super-resolution: Dataset and study
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
math.OC 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Restarted contractive operators to learn at equilibrium
ReTune proves that restarting a contractive truncated unrolled solver and backpropagating only through the final block yields gradients arbitrarily close to the deep-equilibrium hypergradient.