Research map / Probabilistic and scientific modeling / Molecules and proteins
Sequential Latent Knowledge Selection for Knowledge-Grounded Dialogue
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 |
|---|---|---|---|
| use sequential latent knowledge selection instead of the E2E BERT + PostKS + Copy baseline | SKT · Wizard of Wikipedia Test Unseen | knowledge accuracy: 14.1 -> 18.3 | Helped |
| use sequential latent knowledge selection instead of the E2E BERT + PostKS + Copy baseline | SKT · Wizard of Wikipedia Test Unseen | R-1: 15.6 -> 16.1 | Helped |
| add the copy mechanism (ablation) | E2E BERT + PostKS · Wizard of Wikipedia Test Unseen | R-1: 13.4 -> 15.6 | Helped |
| add the copy mechanism (ablation) | E2E BERT + PostKS · Wizard of Wikipedia Test Unseen | knowledge accuracy: 14.1 -> 14.4 | No clear effect |
| add the auxiliary knowledge loss (ablation) | SKT · Wizard of Wikipedia Test Seen | knowledge accuracy: 0.3 -> 26.8 | Helped |
| add the auxiliary knowledge loss (ablation) | SKT · Wizard of Wikipedia Test Unseen | knowledge accuracy: 0.1 -> 18.3 | Helped |
| train with fewer knowledge labels (ablation) | SKT · Wizard of Wikipedia Test Seen | knowledge accuracy: 26.8 -> 25.1; 26.8 -> 22.4; 26.8 -> 21.0 | Hurt |
| train with fewer knowledge labels (ablation) | SKT · Wizard of Wikipedia Test Unseen | knowledge accuracy: 18.3 -> 16.7; 18.3 -> 13.8; 18.3 -> 12.3 | Hurt |
| use BERT-based sentence encoding (ablation) | E2E BERT compared with E2E Transformer MemNet · Wizard of Wikipedia Test Seen | knowledge accuracy: 22.5 -> 23.7 | Helped |
Closest papers by meaning
- Diverse and Faithful Knowledge-Grounded Dialogue Generation via Sequential Posterior Inference (ICML 2023 · similarity 0.92)
- Zero-Resource Knowledge-Grounded Dialogue Generation (NeurIPS 2020 · similarity 0.90)
- KPT: Keyword-Guided Pre-training for Grounded Dialog Generation (AAAI 2023 · similarity 0.89)
- Low-Resource Knowledge-Grounded Dialogue Generation (ICLR 2020 · similarity 0.89)
- Improving the Robustness of Knowledge-Grounded Dialogue via Contrastive Learning (AAAI 2024 · similarity 0.88)
- Improving Open-Domain Dialogue Response Generation with Multi-Source Multilingual Commonsense Knowledge (AAAI 2024 · similarity 0.88)
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.