Research map / Architectures and efficiency / RNNs and normalization
Lattice-Based Recurrent Neural Network Encoders for Neural Machine Translation
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 |
|---|---|---|---|
| Encode source input with a word lattice instead of a single CTB segmentation (ablation) | DWL-GRU + Gating · NIST Chinese-English test sets (ALL) | BLEU: 29.97 -> 32.95 | Helped |
| Encode source input with a word lattice instead of a single PKU segmentation (ablation) | DWL-GRU + Gating · NIST Chinese-English test sets (ALL) | BLEU: 29.68 -> 32.95 | Helped |
| Encode source input with a word lattice instead of a single MSR segmentation (ablation) | DWL-GRU + Gating · NIST Chinese-English test sets (ALL) | BLEU: 28.14 -> 32.95 | Helped |
| Use word-level segmentation rather than character-level input | RNNSearch · NIST Chinese-English test sets (ALL) | BLEU: 30.74 -> 31.78 | Helped |
| Add lattice-based encoding rather than combining multiple tokenizations only at the final time step | LatticeNMT versus MultiSourceNMT · NIST Chinese-English test sets (ALL) | BLEU: 32.13 -> 32.95 | Helped |
| Use DWL-GRU rather than SWL-GRU with pooling | LatticeNMT · NIST Chinese-English test sets (ALL) | BLEU: 32.69 -> 32.71 | Helped |
| Use DWL-GRU rather than SWL-GRU with gating | LatticeNMT · NIST Chinese-English test sets (ALL) | BLEU: 32.79 -> 32.95 | Helped |
| Use gating composition rather than pooling composition (ablation) | DWL-GRU · NIST Chinese-English test sets (ALL) | BLEU: 32.71 -> 32.95 | Helped |
| Use gating composition rather than pooling composition (ablation) | SWL-GRU · NIST Chinese-English test sets (ALL) | BLEU: 32.69 -> 32.79 | Helped |
| Use lattice-based encoding rather than standard RNNSearch with CTB input | DWL-GRU + Gating · NIST Chinese-English test sets (ALL) | BLEU: at least 1.17 BLEU points | Helped |
Closest papers by meaning
- Multi-Channel Encoder for Neural Machine Translation (AAAI 2018 · similarity 0.89)
- From Fully Trained to Fully Random Embeddings: Improving Neural Machine Translation with Compact Word Embedding Tables (AAAI 2022 · similarity 0.89)
- Lattice Recurrent Unit: Improving Convergence and Statistical Efficiency for Sequence Modeling (AAAI 2018 · similarity 0.88)
- Neural Machine Translation with Joint Representation (AAAI 2020 · similarity 0.87)
- Graph Based Translation Memory for Neural Machine Translation (AAAI 2019 · similarity 0.87)
- M$^2$RNN: Non-Linear RNNs with Matrix-Valued States for Scalable Language Modeling (NeurIPS 2026 · 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.