Research map / Image generation / Text-to-image diffusion
Generating Multiple Diverse Responses for Short-Text Conversation
Jun Gao, Wei Bi, Xiaojiang Liu, Junhui Li, Shuming Shi
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 minimum rather than average loss across the response set (ablation) | Ours · Weibo | BLEU-4: 10.086 -> 10.945 | Improved |
| use minimum rather than average loss across the response set (ablation) | Ours · Weibo | %Good: 8.556 -> 20.944 | Improved |
| use minimum rather than average loss across the response set (ablation) | Ours · Twitter | BLEU-4: 34.891 -> 49.965 | Improved |
| use minimum rather than average loss across the response set (ablation) | Ours · Twitter | %Good: 4.444 -> 7.944 | Improved |
| jointly train the pretrained networks with the RL algorithm (ablation) | Ours · Weibo | %Good: 9.611 -> 20.944 | Improved |
| jointly train the pretrained networks with the RL algorithm (ablation) | Ours · Twitter | BLEU-4: 34.282 -> 49.965 | Improved |
| use a large latent-word space rather than a smaller one (ablation) | Ours(Min) · Twitter | BLEU | Improved |
| generate responses using jointly modeled multiple replies and latent words | Ours(Min) · Weibo | %Good: 7.111 -> 20.944 | Improved |
| generate responses using jointly modeled multiple replies and latent words | Ours(Min) · Twitter | %Good: 4.667 -> 7.944 | Improved |
Closest papers by meaning
- SimpleStrat: Diversifying Language Model Generation with Stratification (NeurIPS 2025 · similarity 0.87)
- Generating Informative and Diverse Conversational Responses via Adversarial Information Maximization (NeurIPS 2018 · similarity 0.87)
- Posterior-GAN: Towards Informative and Coherent Response Generation with Posterior Generative Adversarial Network (AAAI 2020 · similarity 0.87)
- AvgOut: A Simple Output-Probability Measure to Eliminate Dull Responses (AAAI 2020 · similarity 0.87)
- Multiresolution Recurrent Neural Networks: An Application to Dialogue Response Generation (AAAI 2017 · similarity 0.87)
- Towards a Neural Conversation Model With Diversity Net Using Determinantal Point Processes (AAAI 2018 · 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.
