fuzhenxin/Style-Transfer-in-Text

Organizes 100+ peer-reviewed papers on text style transfer across supervised and unsupervised approaches, including back-translation, adversarial autoencoders, and reinforcement learning methods. Covers specialized applications like formality conversion, sentiment modification, and toxicity removal, alongside foundational datasets (YAFC, ParaDetox) and evaluation frameworks. Extends beyond style transfer to related controlled text generation and disentangled representation learning techniques.

1,625 stars. No commits in the last 6 months.