Papers by Sascha Rothe
A Thorough Evaluation of Task-Specific Pretraining for Summarization (2021.emnlp-main)
Copied to clipboard
| Challenge: | Previous work has used task-agnostic pretraining methods like masked language models or corrupted span prediction to improve performance on downstream tasks. |
| Approach: | They propose to use a task-agnostic pretraining to improve on low-resource tasks. |
| Outcome: | The proposed model can predict extracted gap sentences on summarization with a low resource and zero shot setup. |
Encode, Tag, Realize: High-Precision Text Editing (D19-1)
Copied to clipboard
| Challenge: | Neural sequence-to-sequence models provide a powerful framework for learning to translate source texts into target texts. |
| Approach: | They propose a sequence tagging approach that casts text generation as a text editing task. |
| Outcome: | The proposed model outperforms strong seq2seq models on sentence fusion, sentence splitting, abstractive summarization, and grammar correction tasks and achieves state-of-the-art performance. |
LM-CPPF: Paraphrasing-Guided Data Augmentation for Contrastive Prompt-Based Few-Shot Fine-Tuning (2023.acl-short)
Copied to clipboard
| Challenge: | Recent advances in pre-trained language models have been limited when fine-tuned on small datasets. |
| Approach: | They propose to add contrastive learning to prompt-based fine-tuning to improve model performance. |
| Outcome: | The proposed approach outperforms other methods on multiple text classification benchmarks. |
On Relation-Specific Neurons in Large Language Models (2025.emnlp-main)
Copied to clipboard
Yihong Liu, Runsheng Chen, Lea Hirlimann, Ahmad Dawar Hakimi, Mingyang Wang, Amir Hossein Kargaran, Sascha Rothe, François Yvon, Hinrich Schuetze
| Challenge: | In large language models, certain neurons can store distinct pieces of knowledge learned during pretraining. |
| Approach: | They hypothesize that relation-specific neurons detect relation in input text and guide generation involving such a relation. |
| Outcome: | The proposed model can handle facts involving relation r and facts containing a different relation . |
Unsupervised Text Style Transfer with Padded Masked Language Models (2020.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for style transfer are difficult to obtain and require substantial amounts of parallel training examples to work well. |
| Approach: | They propose an unsupervised method for style transfer that uses masked language models to find the text spans where the two models disagree the most in terms of likelihood. |
| Outcome: | The proposed method performs competitively in a fully unsupervised setting and improves accuracy in low-resource settings by over 10 percentage points when pre-training on silver training data generated by Masker. |
Focus Attention: Promoting Faithfulness and Diversity in Summarization (2021.acl-long)
Copied to clipboard
| Challenge: | Currently, document summarization is challenging even for humans. |
| Approach: | They propose a focus attention mechanism which encourages decoders to generate tokens that are topically similar to the input document. |
| Outcome: | The proposed method outperforms top-k and nucleus sampling methods on the BBC extreme summarization task and is more accurate than focus attention-based models. |
Leveraging Pre-trained Checkpoints for Sequence Generation Tasks (2020.tacl-1)
Copied to clipboard
| Challenge: | Unsupervised pre-training of large neural models has revolutionized Natural Language Processing. |
| Approach: | They propose to use pre-trained checkpoints for Sequence Generation to initialize a Transformer-based sequence-to-sequence model that is compatible with these checkpoint. |
| Outcome: | The proposed model is compatible with pre-trained BERT, GPT-2, and RoBERTa checkpoints and achieves state-of-the-art results on Machine Translation, Text Summarization, Sentence Splitting, and Sentance Fusion. |
A Simple Recipe for Multilingual Grammatical Error Correction (2021.acl-short)
Copied to clipboard
| Challenge: | Modern approaches view the task of Grammatical Error Correction (GEC) as monolingual text-to-text rewriting and employ encoderdecoder neural architectures. |
| Approach: | They propose a language-agnostic method to generate a large number of synthetic examples and use large-scale multilingual language models to train state-of-the-art GEC models. |
| Outcome: | The proposed method surpasses state-of-the-art results on GEC benchmarks in English, Czech, German and Russian. |