atlas

Research map / Large language models / Retrieval and knowledge

Zero-Resource Knowledge-Grounded Dialogue Generation

NeurIPS 2020 · Large language models › Retrieval and knowledge · approach: Explanatory and multiple-choice QA benchmarks · 56 citations (OpenAlex)

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 changedWhereMeasuredEffect
Use zero-resource variational learning with dialogue and knowledge corpora that are independentZRKGC · Wizard Seen and Wizard UnseenF1: 18.7 vs 19.3; 18.6 vs 17.9Mixed
Use zero-resource variational learning with independent dialogue and knowledge corporaZRKGC · Topical Freq, Topical Rare, and CMU_DoGF1: 16.6 vs 16.5; 16.8 vs 15.1; 12.5 vs 10.7Helped
Add the grounding-rate latent variable Zα (ablation)ZRKGC · Wizard SeenF1: 18.5 -> 18.7Helped
Add the mutual information loss (ablation)ZRKGC · Wizard SeenF1: 18.1 -> 18.7Helped
Use a retrieval posterior instead of a generative posterior (ablation)ZRKGC · Wizard SeenF1: 16.2 -> 18.7Helped
Parameterize the posterior over retrieved knowledge candidates rather than use only the top candidate (ablation)ZRKGC · Wizard SeenF1: 17.2 -> 18.7Helped
Use the learned knowledge selection module rather than remove it (ablation)ZRKGC · Wizard SeenF1: 18.3 -> 18.7Helped

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy