Research map / Probabilistic and scientific modeling / Variational and generative sampling
Variational Memory Encoder-Decoder
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 VMED (K=4) rather than VLSTM | Cornell Movies response generation · Cornell Movies | BLEU-4: 9.7 -> 12.3 | Helped |
| Use VMED (K=3) rather than VLSTM | OpenSubtitle response generation · OpenSubtitle | BLEU-4: 8.1 -> 12.9 | Helped |
| Use VMED (K=3) rather than Seq2Seq | LJ users question-answering response generation · LJ users question-answering | BLEU-4: 6.4 -> 9.8 | Helped |
| Use VMED (K=3) rather than DNC | Reddit comments response generation · Reddit comments | BLEU-4: 3.4 -> 6.4 | Helped |
| Use K=4 instead of K=1 (ablation) | VMED · Cornell Movies | BLEU-4: 10.8 -> 12.3 | Helped |
| Use K=4 instead of K=2 (ablation) | VMED · Cornell Movies | BLEU-4: 11.9 -> 12.3 | Helped |
| Use K=4 instead of K=3 (ablation) | VMED · Cornell Movies | BLEU-4: 10.4 -> 12.3 | Helped |
| Use K=3 instead of K=1 (ablation) | VMED · OpenSubtitle | BLEU-4: 6.2 -> 12.9 | Helped |
| Use K=3 instead of K=2 (ablation) | VMED · OpenSubtitle | BLEU-4: 8.8 -> 12.9 | Helped |
| Use K=3 instead of K=4 (ablation) | VMED · OpenSubtitle | BLEU-4: 9.3 -> 12.9 | Helped |
| Use K=3 instead of K=1 (ablation) | VMED · LJ users question-answering | BLEU-4: 6.9 -> 9.8 | Helped |
| Use K=3 instead of K=2 (ablation) | VMED · LJ users question-answering | BLEU-4: 7.9 -> 9.8 | Helped |
Closest papers by meaning
- Improving Variational Encoder-Decoders in Dialogue Generation (AAAI 2018 · similarity 0.91)
- A Variational Framework for Improving Naturalness in Generative Spoken Language Models (ICML 2025 · similarity 0.85)
- Memory-Augmenting Decoder-Only Language Models through Encoders (Student Abstract) (AAAI 2024 · similarity 0.85)
- Beyond Vanilla Variational Autoencoders: Detecting Posterior Collapse in Conditional and Hierarchical Variational Autoencoders (ICLR 2024 · similarity 0.85)
- Memory Decoder: A Pretrained, Plug-and-Play Memory for Large Language Models (NeurIPS 2025 · similarity 0.85)
- MBR and QE Finetuning: Training-time Distillation of the Best and Most Expensive Decoding Methods (ICLR 2024 · 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.