Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Learning to Know Myself: A Coarse-to-Fine Persona-Aware Training Framework for Personalized Dialogue Generation
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 |
|---|---|---|---|
| Apply coarse-to-fine persona-aware training | DialoGPT · ConvAI2 validation set | C: 55.19 -> 63.90 | Helped |
| Apply coarse-to-fine persona-aware training | GPT2 · Persona-Chat testing set | C: 47.86 -> 62.22 | Helped |
| Apply coarse-to-fine persona-aware training | GPT2 · DuLeMon testing set, full data | C: 64.12 -> 69.57 | Helped |
| Add coarse-grained persona-aware training K (ablation) | GPT2 · Persona-Chat testing set, Original Persona | C: 47.86 -> 52.70 | Helped |
| Add fine-grained contrastive training A (ablation) | GPT2 · Persona-Chat testing set, Original Persona | C: 47.86 -> 55.71 | Helped |
| Add both coarse-grained training K and contrastive training A (ablation) | GPT2 · Persona-Chat testing set, Original Persona | C: 47.86 -> 62.22 | Helped |
| Add coarse-grained persona-aware training K (ablation) | Per-Seq2Seq · Persona-Chat testing set, Original Persona | C: 22.96 -> 35.69 | Helped |
| Add fine-grained contrastive training A (ablation) | Per-Seq2Seq · Persona-Chat testing set, Original Persona | C: 22.96 -> 31.39 | Helped |
| Add both coarse-grained training K and contrastive training A (ablation) | Per-Seq2Seq · Persona-Chat testing set, Original Persona | C: 22.96 -> 41.72 | Helped |
| Add coarse-grained persona-aware training K (ablation) | GPT2 · Persona-Chat testing set, Revised Persona | C: 6.04 -> 22.98 | Helped |
| Add fine-grained contrastive training A (ablation) | GPT2 · Persona-Chat testing set, Revised Persona | C: 6.04 -> 21.99 | Helped |
| Add both coarse-grained training K and contrastive training A (ablation) | GPT2 · Persona-Chat testing set, Revised Persona | C: 6.04 -> 31.23 | Helped |
Closest papers by meaning
- A Disentangled-Attention Based Framework with Persona-Aware Prompt Learning for Dialogue Generation (AAAI 2023 · similarity 0.94)
- Personalized Dialogue Generation with Persona-Adaptive Attention (AAAI 2023 · similarity 0.93)
- Consistently Simulating Human Personas with Multi-Turn Reinforcement Learning (NeurIPS 2025 · similarity 0.87)
- Enhancing Personalized Multi-Turn Dialogue with Curiosity Reward (NeurIPS 2025 · similarity 0.86)
- PersonaVLM: Long-Term Personalized Multimodal LLMs (CVPR 2026 · similarity 0.85)
- Teaching Language Models to Evolve with Users: Dynamic Profile Modeling for Personalized Alignment (NeurIPS 2025 · similarity 0.84)
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.