Research map / Label-efficient and robust learning / Recognition and long-tail
RefNet: A Reference-Aware Network for Background Based Conversation
Chuan Meng, Pengjie Ren, Zhumin Chen, Christof Monz, Jun Ma, Maarten de Rijke
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 |
|---|---|---|---|
| Combine reference decoding with generation decoding (ablation) | RefNet · Holl-E mixed-short | BLEU: 26.73 -> 30.33 (SR); 26.01 -> 30.33 (SR); 30.84 -> 33.97 (MR); 29.19 -> 33.97 (MR) | Improved |
| Add supervised switcher loss (ablation) | RefNet · Holl-E mixed-short, SR | F1: 37.13 -> 41.86 | Improved |
| Add supervised switcher loss (ablation) | RefNet · Holl-E mixed-short, MR | BLEU: 31.63 -> 33.97 | Improved |
| Use RefNet instead of CaKe | Holl-E mixed-short · Holl-E mixed-short, SR | BLEU: 26.03 -> 30.33 | Improved |
| Use RefNet instead of QANet | Holl-E mixed-short · Holl-E mixed-short, SR | ROUGE-1: 40.66 -> 42.11 | Improved |
| Use RefNet instead of CaKe | Holl-E mixed-short, human evaluation · Holl-E mixed-short | Naturalness (at least 2 workers): 264 -> 299 | Improved |
| Use RefNet instead of QANet | Holl-E mixed-short, human evaluation · Holl-E mixed-short | Informativeness (at least 2 workers): 225 -> 247 | Improved |
Closest papers by meaning
- Thinking Globally, Acting Locally: Distantly Supervised Global-to-Local Knowledge Selection for Background Based Conversation (AAAI 2020 · similarity 0.87)
- Breaking the Reference Bottleneck via Learning to Rewrite Conversational Queries without Gold Reference Passages (ICML 2026 · similarity 0.86)
- Dynamic Knowledge Routing Network for Target-Guided Open-Domain Conversation (AAAI 2020 · similarity 0.85)
- ChatterBox: Multimodal Referring and Grounding with Chain-of-Questions (AAAI 2025 · similarity 0.84)
- RefDetector: A Simple Yet Effective Matching-based Method for Referring Expression Comprehension (AAAI 2025 · similarity 0.83)
- SAMA: Towards Multi-Turn Referential Grounded Video Chat with Large Language Models (NeurIPS 2025 · 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.
