Research map / Large language models / Retrieval and knowledge
What Do You Mean ‘Why?’: Resolving Sluices in Conversations
Victor Petrén Bach Hansen, Anders Søgaard
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| Fine-tune GPT-2 on the sluicing corpus rather than use pretrained GPT-2 unchanged (ablation) | GPT-2 · Conversational sluicing test | GLEU: 0.067 -> 0.348 | Improved |
| Fine-tune GPT-2 on the sluicing corpus rather than use pretrained GPT-2 unchanged (ablation) | GPT-2 · Conversational sluicing test | BLEU: 0.117 -> 0.391 | Improved |
| Fine-tune GPT-2 on the sluicing corpus rather than use pretrained GPT-2 unchanged (ablation) | GPT-2 · Conversational sluicing test | chrF: 0.138 -> 0.467 | Improved |
| Fine-tune GPT-2 on the sluicing corpus rather than use pretrained GPT-2 unchanged (ablation) | GPT-2 · Conversational sluicing test | chrP: 0.109 -> 0.499 | Improved |
| Fine-tune GPT-2 on the sluicing corpus rather than use pretrained GPT-2 unchanged (ablation) | GPT-2 · Conversational sluicing test | chrR: 0.167 -> 0.470 | Improved |
| Pretrain LSTM-seq2seq and Transformer with denoising autoencoder objectives before sluicing-task fine-tuning (ablation) | LSTM-seq2seq and Transformer · Conversational sluicing corpus | Automatic metrics | No clear effect |
| Replace a one-word sluice with its gold annotated resolution in CoQA development inputs | CoQA baseline QA model · CoQA development | Answer quality | No clear effect |
Closest papers by meaning
- Filling Conversation Ellipsis for Better Social Dialog Understanding (AAAI 2020 · similarity 0.83)
- NoLiMa: Long-Context Evaluation Beyond Literal Matching (ICML 2025 · similarity 0.81)
- Bridging Context Gaps: Leveraging Coreference Resolution for Long Contextual Understanding (ICLR 2025 · similarity 0.81)
- Mutable Transcripts: Mitigating Context Pollution through Editable Conversation State (NeurIPS 2026 · similarity 0.81)
- Introducing Visual Scenes and Reasoning: A More Realistic Benchmark for Spoken Language Understanding (AAAI 2026 · similarity 0.81)
- ChatterBox: Multimodal Referring and Grounding with Chain-of-Questions (AAAI 2025 · similarity 0.81)
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.
