Research map / Image generation / Autoregressive and latent generation
Creativity: Generating Diverse Questions Using Variational Autoencoders
Check 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 exploratory uniform latent sampling instead of normal sampling (ablation) | VAE-LSTM on VQG-COCO · VQG-COCO | BLEU oracle: 0.393 -> 0.579 | Helped |
| Use exploratory uniform latent sampling instead of normal sampling (ablation) | VAE-LSTM on VQG-COCO · VQG-COCO | BLEU average: 0.356 -> 0.311 | Hurt |
| Use exploratory uniform latent sampling instead of normal sampling (ablation) | VAE-LSTM on VQG-COCO · VQG-COCO | METEOR oracle: 0.219 -> 0.342 | Helped |
| Use exploratory uniform latent sampling instead of normal sampling (ablation) | VAE-LSTM on VQG-COCO · VQG-COCO | Generative strength: 1.98 -> 46.10 | Helped |
| Use exploratory uniform latent sampling instead of normal sampling (ablation) | VAE-LSTM on VQG-COCO · VQG-COCO | Inventiveness: 10.76 -> 27.88 | Helped |
| Increase latent samples from 100 to 500 with normal sampling (ablation) | VAE-LSTM on VQG-COCO · VQG-COCO | BLEU average: 0.356 -> 0.352 | Hurt |
| Increase latent samples from 100 to 500 with U20 sampling (ablation) | VAE-LSTM on VQG-COCO · VQG-COCO | BLEU oracle: 0.544 -> 0.579 | Helped |
| Increase latent samples from 100 to 500 with U20 sampling (ablation) | VAE-LSTM on VQG-COCO · VQG-COCO | Generative strength: 22.01 -> 46.10 | Helped |
Closest papers by meaning
- Information Maximizing Visual Question Generation (CVPR 2019 · similarity 0.88)
- Learning to Generate Visual Questions with Noisy Supervision (NeurIPS 2021 · similarity 0.86)
- Automated Generation of Challenging Multiple-Choice Questions for Vision Language Model Evaluation (CVPR 2025 · similarity 0.85)
- ConVQG: Contrastive Visual Question Generation with Multimodal Guidance (AAAI 2024 · similarity 0.84)
- Yo'Chameleon: Personalized Vision and Language Generation (CVPR 2025 · similarity 0.83)
- Explicitly Guided Difficulty-Controllable Visual Question Generation (AAAI 2025 · similarity 0.83)
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.