Research map / Large language models / Retrieval and knowledge
Reasoning in Dialog: Improving Response Generation by Context Reading Comprehension
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 |
|---|---|---|---|
| jointly train response generation and reading comprehension instead of response generation alone (ablation) | MRG · DRCD response generation test | BLEU4: 0.0661 -> 0.0741 | Helped |
| jointly train response generation and reading comprehension instead of reading comprehension alone (ablation) | MRG · DRCD reading comprehension test | Mean accuracy (%): 44.89 -> 45.43 | Helped |
| add memory-augmented cross-attention (MCAM) (ablation) | MRG · DRCD response generation test | BLEU1: 0.2224 -> 0.2632 | Helped |
| add hierarchical self-attention (MAM) (ablation) | MRG · DRCD response generation test | BLEU1: 0.2404 -> 0.2632 | Helped |
| add the memory updater (ablation) | MRG · DRCD response generation test | BLEU1: 0.2498 -> 0.2632 | Helped |
| add the memory updater (ablation) | MRG · DRCD response generation test | BLEU4: 0.0747 -> 0.0741 | No clear effect |
| replace PAC with MRG | dialog response generation · DRCD response generation test | BLEU1: 0.2413 -> 0.2632 | Helped |
| use MRG rather than the strongest baseline, DMN+ | dialog reading comprehension · DRCD reading comprehension test | Mean accuracy (%): 43.97 -> 45.43 | Helped |
Closest papers by meaning
- Local Explanation of Dialogue Response Generation (NeurIPS 2021 · similarity 0.91)
- Dialog-based Language Learning (NeurIPS 2016 · similarity 0.89)
- Dialog-to-Action: Conversational Question Answering Over a Large-Scale Knowledge Base (NeurIPS 2018 · similarity 0.89)
- TA&AT: Enhancing Task-Oriented Dialog with Turn-Level Auxiliary Tasks and Action-Tree Based Scheduled Sampling (AAAI 2024 · similarity 0.88)
- From Retrieval to Generation: A Simple and Unified Generative Model for End-to-End Task-Oriented Dialogue (AAAI 2024 · similarity 0.88)
- StateTree: Enhancing Long-term Dialogue Reasoning via Reinforcement Learning (NeurIPS 2026 · similarity 0.86)
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.