Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Dual Task Framework for Improving Persona-Grounded Dialogue Dataset
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 Persona-Link with the large PKB instead of no persona augmentation | Persona-Chat Bi-encoder · Persona-Chat testset | R@1/20: 0.814 -> 0.931 | Helped |
| Increase the Persona-Link PKB from 1K to 5K personas (ablation) | Persona-Link · Persona-Chat testset | R@1/20: 0.904 -> 0.931 | Helped |
| Augment personas with Persona-Link at test time | Wikipedia-pretrained dialogue model with no persona · Persona-Chat testset, no persona | R@1/20: 0.591 -> 0.733 | Helped |
| Augment personas with Persona-Link at test time | Reddit-pretrained dialogue model with no persona · Persona-Chat testset, no persona | R@1/20: 0.659 -> 0.805 | Helped |
| Augment personas with Persona-Link at test time | Wikipedia-pretrained dialogue model with incomplete persona · Persona-Chat testset, incomplete persona | R@1/20: 0.768 -> 0.785 | Helped |
| Augment personas with Persona-Link at test time | Reddit-pretrained dialogue model with incomplete persona · Persona-Chat testset, incomplete persona | R@1/20: 0.825 -> 0.858 | Helped |
| Use commonsense expansion and label regularization instead of the semantic-only Bi-encoder (ablation) | Persona-Link · Persona-Link | R@1: 0.583 -> 0.669 | Helped |
| Replace commonsense expansion with EDA (ablation) | Persona-Link · Persona-Link | R@1: 0.669 -> 0.599 | Hurt |
| Replace commonsense expansion with paraphrasing (ablation) | Persona-Link · Persona-Link | R@1: 0.669 -> 0.610 | Hurt |
| Use the proposed Persona-Link rather than a Bi-encoder linker | Dialogue models with inference-time persona augmentation · Persona-Chat testset, no persona | R@1/20: 0.699 -> 0.733; 0.773 -> 0.805 | Helped |
| Use the proposed Persona-Link rather than a Bi-encoder linker | Dialogue models with inference-time persona augmentation · Persona-Chat testset, incomplete persona | R@1/20: 0.768 -> 0.785; 0.845 -> 0.858 | Helped |
| Use out-dialogue semantic matching rather than in-dialogue matching | Persona-Link training data · Persona-Link alignments | linguistic similarity: 0.2 to 0.1 | Helped |
Closest papers by meaning
- Learning to Know Myself: A Coarse-to-Fine Persona-Aware Training Framework for Personalized Dialogue Generation (AAAI 2023 · similarity 0.92)
- A Pre-Training Based Personalized Dialogue Generation Model with Persona-Sparse Data (AAAI 2020 · similarity 0.92)
- Personalized Dialogue Generation with Persona-Adaptive Attention (AAAI 2023 · similarity 0.92)
- Consistently Simulating Human Personas with Multi-Turn Reinforcement Learning (NeurIPS 2025 · similarity 0.86)
- Teaching Language Models to Evolve with Users: Dynamic Profile Modeling for Personalized Alignment (NeurIPS 2025 · similarity 0.85)
- Enhancing Personalized Multi-Turn Dialogue with Curiosity Reward (NeurIPS 2025 · similarity 0.85)
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.