Research map / Architectures and efficiency / Pruning and efficient tuning
Monolingual Transfer Learning via Bilingual Translators for Style-Sensitive Paraphrase Generation
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add roundtrip-translation pre-training (RT) before fine-tuning (ablation) | RNN paraphrase generator · Informal->Formal F&R | BLEU: 67.38 -> 74.28 vs AE; Base 66.45 -> RT 69.03 on E&M | Helped |
| add AutoEncoder pre-training before fine-tuning (ablation) | paraphrase generators (RNN/CNN/SAN) · GYAFC (4 tasks) | BLEU: 57.76 -> 67.38 (RNN F&R I->F); 24.85/30.93 CNN-Base vs 28.12 CNN-AE F->I E&M | Mixed |
| use RT pre-training instead of AE pre-training (ablation) | SAN · Informal->Formal E&M | BLEU: 65.57 -> 69.58 | Helped |
| use RT pre-training instead of back-translation pre-training (ablation) | SAN · Informal->Formal F&R | BLEU: 73.39 -> 74.19 | Helped |
| add back-translation pre-training over AE (ablation) | SAN · Informal->Formal E&M | BLEU: 65.57 -> 69.06 | Helped |
| ensemble RNN-RT vs previous ensemble MultiTask | formality style transfer · Informal->Formal E&M | BLEU: 72.13 -> 72.41 | Helped |
| RT pre-training for text simplification | SAN · WikiSmall | SARI: 34.19 -> 35.46 | Helped |
| RT pre-training for text simplification | SAN · WikiLarge | SARI: 34.58 -> 35.99 | Helped |
| RT pre-training with reduced fine-tuning data (5k pairs) | SAN · Informal->Formal E&M/F&R | BLEU | Helped |
Closest papers by meaning
- A Probabilistic Formulation of Unsupervised Text Style Transfer (ICLR 2020 · similarity 0.88)
- ParaGuide: Guided Diffusion Paraphrasers for Plug-and-Play Textual Style Transfer (AAAI 2024 · similarity 0.88)
- PARABANK: Monolingual Bitext Generation and Sentential Paraphrasing via Lexically-Constrained Neural Machine Translation (AAAI 2019 · similarity 0.88)
- Adaptive Prompt Routing for Arbitrary Text Style Transfer with Pre-trained Language Models (AAAI 2024 · similarity 0.85)
- LAMPAT: Low-Rank Adaption for Multilingual Paraphrasing Using Adversarial Training (AAAI 2024 · similarity 0.85)
- A Paradigm Shift in Machine Translation: Boosting Translation Performance of Large Language Models (ICLR 2024 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.