An effective-sample-size-driven annealing schedule trains normalizing flows without mode collapse and estimates the repressilator model's marginal likelihood roughly ten times faster than ensemble MCMC.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Mitigating mode collapse in normalizing flows by annealing with an adaptive schedule: Application to parameter estimation
An effective-sample-size-driven annealing schedule trains normalizing flows without mode collapse and estimates the repressilator model's marginal likelihood roughly ten times faster than ensemble MCMC.