Pith. sign in

Generative Knowledge Transfer for Neural Language Models

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it
abstract

In this paper, we propose a generative knowledge transfer technique that trains an RNN based language model (student network) using text and output probabilities generated from a previously trained RNN (teacher network). The text generation can be conducted by either the teacher or the student network. We can also improve the performance by taking the ensemble of soft labels obtained from multiple teacher networks. This method can be used for privacy conscious language model adaptation because no user data is directly used for training. Especially, when the soft labels of multiple devices are aggregated via a trusted third party, we can expect very strong privacy protection.

fields

cs.CV 1

years

2024 1

verdicts

CONDITIONAL 1

representative citing papers

FlexPose: Pose Distribution Adaptation with Limited Guidance

cs.CV · 2024-12-18 · conditional · novelty 6.0

A few-shot method that transfers a source pose distribution to a target one by fine-tuning only a few linear layers of a StyleGAN skeleton generator, with pose-mixup and sparse regularization.

citing papers explorer

Showing 1 of 1 citing paper.

  • FlexPose: Pose Distribution Adaptation with Limited Guidance cs.CV · 2024-12-18 · conditional · none · ref 38 · internal anchor

    A few-shot method that transfers a source pose distribution to a target one by fine-tuning only a few linear layers of a StyleGAN skeleton generator, with pose-mixup and sparse regularization.