Research map / Label-efficient and robust learning / Knowledge distillation
Distillation for machine translation
Applies selective and uncertainty-aware distillation to non-autoregressive and simultaneous translation.
18 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2019–2026) · an approach in Knowledge distillation, keyword cluster “distilled · ensemble · synthetic”
State History Tried Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
11 papers in 2023–24 and 2 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic shrank, from 7.7% to 0.9%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2019 · Multilingual Neural Machine Translation with Knowledge Distillation (ICLR 2019 · 127 citations)
- 2021 · FastSpeech 2: Fast and High-Quality End-to-End Text to Speech (ICLR 2021 · 506 citations)
- 2022 · Measuring and Reducing Model Update Regression in Structured Prediction for NLP (NeurIPS 2022 · 3 citations)
- 2023 · Cross-Modal Distillation for Speaker Recognition (AAAI 2023 · 21 citations)
- 2024 · Discrepancy and Uncertainty Aware Denoising Knowledge Distillation for Zero-Shot Cross-Lingual Named Entity Recognition (AAAI 2024 · 8 citations)
- 2025 · Implicit Word Reordering with Knowledge Distillation for Cross-Lingual Dependency Parsing (AAAI 2025 · 1 citation)
- 2026 · Entropy-aware Span-Constrained Optimal Transport for Robust Cross-Tokenizer Knowledge Distillation (ICML 2026)
Earliest papers here
- Multilingual Neural Machine Translation with Knowledge Distillation (ICLR 2019 · 127 citations)
- FastSpeech 2: Fast and High-Quality End-to-End Text to Speech (ICLR 2021 · 506 citations)
- Understanding and Improving Lexical Choice in Non-Autoregressive Translation (ICLR 2021 · 55 citations)
What has been tried
17 papers here report 214 measured changes (a component added, removed or swapped, compared with a baseline): helped 77%, no clear effect 4%, hurt 15%, mixed 3%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| add discrepancy-aware representation learning | Wikiann target-language Avg (Table 2) · token-level F1 (%) | 67.21 -> 70.12; ↑2.91 | Helped | Discrepancy and Uncertainty Aware Denoising Knowledge Distillation for Zero-Shot Cross-Lingual Named Entity Recognition (AAAI 2024 · 8 citations) |
| add knowledge distillation from the teacher model | FG-RECCON · EM | 30.49 -> 33.85 | Helped | From Coarse to Fine: A Distillation Method for Fine-Grained Emotion-Causal Span Pair Extraction in Conversation (AAAI 2024 · 3 citations) |
| convert soft word-order distributions to hard pseudo-labels | UD v2.14, average over all languages · UAS | 74.1 -> 73.5 | Hurt | Implicit Word Reordering with Knowledge Distillation for Cross-Lingual Dependency Parsing (AAAI 2025 · 1 citation) |
| Add RNNG distillation | PTB test · F1 | 70.5 -> 69.7 | Hurt | Ensemble Distillation for Unsupervised Constituency Parsing (ICLR 2024 · 0 citations) |
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 13 papers; from OpenAlex and the papers' affiliations).
- Beihang University · 2 papers
Use it
Download BibTeX of 18 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Patient consistent distillation (68 papers)
- Self and data-free distillation (57 papers)
Elsewhere in the topic
- Rethinking distillation objectives (160 papers)
- Distillation with causal and adaptive guidance (82 papers)
- Few-shot and metric distillation (80 papers)
- Teacher properties and robustness (54 papers)
- Knowledge distillation for detectors (25 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).