The DOI matches its Crossref record; 57 of the 57 references it cites were found in a registry or an index; 4 of the 9 sources its claims cite were checked for support.
Sequence-Level Knowledge Distillation
Open accessAbstract
Neural machine translation (NMT) offers a novel alternative formulation of translation that is potentially simpler than statistical approaches.However to reach competitive performance, NMT models need to be exceedingly large.In this paper we consider applying knowledge distillation approaches (Bucila et al., 2006;Hinton et al., 2015) that have proven successful for reducing the size of neural models in other domains to the problem of NMT.We demonstrate that standard knowledge distillation applied to word-level prediction can be effective for NMT, and also introduce two novel sequence-level versions of knowledge distillation that further improve performance, and somewhat surprisingly, seem to eliminate the need for beam search (even when applied on the original teacher model).Our best student model runs 10 times faster than its state-of-the-art teacher with little loss in performance.It is also significantly better than a baseline model trained without knowledge distillation: by 4.2/1.7 BLEU with greedy decoding/beam search.Applying weight pruning on top of knowledge distillation results in a student model that has 13× fewer parameters than the original teacher model, with a decrease of 0.4 BLEU.
Overview
AI snapshot
Written from the paper’s full text.
TL;DR
The paper applies knowledge distillation to neural machine translation, showing word-level distillation helps and introducing sequence-level variants (Seq-KD and Seq-Inter) that train a smaller student on teacher beam-search outputs. The best student runs 10× faster than its teacher with little BLEU loss, and pruning on top yields 13× fewer parameters with a 0.4 BLEU drop.
Key findings
- Word-level knowledge distillation improves over the baseline, but sequence-level knowledge distillation does better on English→German and similarly on Thai→English; combining them gives further gains for the 2×300 and 2×100 models, though not for the 2×500 model. For English→German 2×500, baseline BLEU is 14.7/17.6 (greedy/beam), Seq-KD is 18.9/19.0, and Seq-KD + Seq-Inter is 18.9/19.3, i.e. +4.2/+1.7 BLEU over baseline.
- Sequence-level distillation reduces the need for beam search: greedy decoding with a fine-tuned teacher (Baseline + Seq-Inter) reaches 19.6 BLEU versus 19.5 for beam search with the original teacher. Seq-KD models also assign much more mass to the approximate mode: 16.9% for the 2×500 English→German Seq-KD model versus 0.9% for the baseline.
- The student model can run 10× faster with greedy decoding than the teacher with beam search on GPU: 1051.3 vs 101.9 source words/sec.
- Weight pruning on top of distillation is orthogonal: pruning 80% of weights in the 2×500 Seq-KD + Seq-Inter student gives 13× fewer parameters than the original teacher with only 0.4 BLEU decrease; pruning 90% gives 8m parameters (26× fewer) with a 1.0 BLEU decrease.
Methods
- Uses an attentional encoder-decoder NMT architecture from Luong et al. (2015). Two settings: English→German WMT 2014 (4m sentences, top 50k vocabulary, teacher 4×1000 LSTM, students 2×300 and 2×500) and Thai→English IWSLT 2015 (90k sentences, 25k vocabulary, teacher 2×500, student 2×100). Evaluation is tokenized BLEU with multi-bleu.perl.
- Three distillation variants: Word-KD (cross-entropy with teacher word distribution plus NLL; α tested in {0.5, 0.9}, with 0.5 better); Seq-KD (student trained on teacher beam-search output, beam K=5); and Seq-Inter (student trained on the teacher beam hypothesis with highest smoothed sentence BLEU to gold, beam K=35, fine-tuned from a pretrained model with learning rate 0.1; English→German Seq-Inter data generated on ~50% of the training set). Methods can be combined.
- Weight pruning removes weights with lowest absolute values, then retrains on Seq-KD data with learning rate 0.2 and fine-tunes toward Seq-Inter data with learning rate 0.1. Speed is measured on GPU (GeForce GTX Titan X), CPU, and smartphone (Samsung Galaxy 6).
Limitations
- Stated: the mode approximation for sequence-level distillation is poor because it approximates an exponentially sized distribution with a single sample; Seq-Inter’s two-sequence interpolation is nonideal because it doubles training data and requires training on both teacher-generated and true sequences conditioned on the same source, which are often different. The paper also calls its illustrative noise function “simple (unrealistic).”
- Stated: experiments cover only English→German and Thai→English NMT; the 4×1000 model could not be opened on the smartphone; larger beam sizes than K=5 did not improve results; and English→German Seq-Inter data was generated on only ~50% of the training set for efficiency.
- Evident from the text: no statistical significance tests are reported, hyperparameter searches are narrow (τ ∈ {1, 1.5, 2}; α ∈ {0.5, 0.9}), and compression is compared mainly against pruning rather than other compression families.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
26 Own result · 13 Own interpretation · 9 No basis given · 3 Sources not found · 2 Partly supported · 1 Supported · 1 Supported so far · accuracy not yet measured
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“While both simple and surprisingly accurate, NMT systems typically need to have very high capacity in order to perform well: Sutskever et al. (2014) used a 4-layer LSTM with 1000 hidden units per layer (herein 4×1000) and Zhou et al. (2016) obtained state-of-the-art results on English → French with a 16-layer LSTM with 512 units per layer.”
§ Sequence-Level Knowledge Distillation > 1 Introduction
What the sources say
- Not checked — no full text of the sourcecites Sutskever et al. (2014) source
- Supportedcites Zhou et al. (2016) source
The Model settings section states that every LSTM layer has 512 memory cells. The Introduction and Conclusion report a depth-16 model that achieved state-of-the-art results on WMT’14 English-to-French, supporting the specific example attributed to Zhou et al. (LLM, conf=0.93)
plaincorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Our best student model runs 10 times faster than its state-of-the-art teacher with little loss in performance.”
§ Sequence-Level Knowledge Distillation > Abstract
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Sequence-level interpolation (Seq-Inter), in addition to improving models trained via Word-KD and Seq-KD, also improves upon the original teacher model that was trained on the actual data but finetuned towards Seq-Inter data (Baseline + Seq-Inter).”
§ Sequence-Level Knowledge Distillation > 5 Results and Discussion
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In fact, greedy decoding with this fine-tuned model has similar performance (19.6) as beam search with the original model (19.5), allowing for faster decoding even with an identically-sized model.”
§ Sequence-Level Knowledge Distillation > 5 Results and Discussion
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Our results suggest that this is indeed the case: the probability mass that Seq-KD models assign to the approximate mode is much higher than is the case for baseline models trained on original data (Table 1: p(t = ŷ)).”
§ Sequence-Level Knowledge Distillation > 5 Results and Discussion
hedgedcorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Sequence-level interpolation (Seq-Inter), in addition to improving models trained via Word-KD and Seq-KD, also improves upon the original teacher model that was trained on the actual data but finetuned towards Seq-Inter data (Baseline + Seq-Inter).”
§ Sequence-Level Knowledge Distillation > 5 Results and Discussion
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We find that while word-level knowledge distillation (Word-KD) does improve upon the baseline, sequence-level knowledge distillation (Seq-KD) does better on English → German and performs similarly on Thai → English.”
§ Sequence-Level Knowledge Distillation > 5 Results and Discussion
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We find that while word-level knowledge distillation (Word-KD) does improve upon the baseline, sequence-level knowledge distillation (Seq-KD) does better on English → German and performs similarly on Thai → English.”
§ Sequence-Level Knowledge Distillation > 5 Results and Discussion
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Seq-Inter offers a compromise between the two, with the greedily-decoded sequence accounting for 7.6% of the distribution.”
§ Sequence-Level Knowledge Distillation > 5 Results and Discussion
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We find that while word-level knowledge distillation (Word-KD) does improve upon the baseline, sequence-level knowledge distillation (Seq-KD) does better on English → German and performs similarly on Thai → English.”
§ Sequence-Level Knowledge Distillation > 5 Results and Discussion
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- It may state two separate claims as one. 15
- Probably not a substantial claim. 6
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Researchers have particularly noted that large networks seem to be necessary for training, but learn redundant representations in the process (Denil et al., 2013).”
§ Sequence-Level Knowledge Distillation > 1 Introduction
What the sources say
- Partly supportedcites Denil et al., 2013 sourceExperimental — accuracy not yet measured
The Abstract (p. 1) reports significant redundancy in model parameters, and the Introduction (p. 2) explains that individual feature weights can be redundant. The paper does not say that large networks are necessary for training; its finding concerns redundant parameterization, not necessarily redundant representations. (LLM, conf=0.85)
hedgeddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“In this paper we consider applying knowledge distillation approaches (Bucila et al., 2006;Hinton et al., 2015) that have proven successful for reducing the size of neural models in other domains to the problem of NMT.”
§ Sequence-Level Knowledge Distillation > Abstract
What the sources say
- Not matchedcites Bucila et al., 2006
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Supportedcites Hinton et al., 2015 source
The Discussion (p. 8) says distillation worked well for transferring knowledge from an ensemble or large model into a smaller model, including in MNIST and acoustic-model experiments. These are successful applications outside NMT, as the citing sentence claims. (LLM, conf=0.94)
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
- Not matched
Every source the paper cites for this claim was checked and supports it.
In the paper
“Training on q(y | x; θ T ) is attractive since it gives more information about other classes for a given data point (e.g.similarity between classes) and has less variance in gradients (Hinton et al., 2015).”
§ Sequence-Level Knowledge Distillation > 2.2 Knowledge Distillation
What the sources say
- Supportedcites Hinton et al., 2015 source
The Introduction (p. 2) says that soft targets from a cumbersome model provide more information per training case and less variance in gradients when they have high entropy. It also explains that probabilities assigned to other classes convey a similarity structure. (LLM, conf=0.93)
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Supported
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Applying weight pruning on top of knowledge distillation results in a student model that has 13× fewer parameters than the original teacher model, with a decrease of 0.4 BLEU.”
§ Sequence-Level Knowledge Distillation > Abstract
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Combining them (Seq-KD + Word-KD) results in further gains for the 2 × 300 and 2 × 100 models (although not for the 2 × 500 model)”
§ Sequence-Level Knowledge Distillation > 5 Results and Discussion
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“A beam of size 5 captures 2.8% of the distribution for German → English, and 3.8% for Thai → English.”
§ Sequence-Level Knowledge Distillation > 3.2 Sequence-Level Knowledge Distillation
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We found this to give comparable results to standard knowledge distillation and hence did not pursue this further.”
§ Sequence-Level Knowledge Distillation > 5.3 Further Observations
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In fact, in experiments we find that with beam of size 1, q(ŷ | s) (on average) accounts for 1.3% of the distribution for German → English, and 2.3% for Thai → English (Table 1: p(t = ŷ)).”
§ Sequence-Level Knowledge Distillation > 3.2 Sequence-Level Knowledge Distillation
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“The perplexity of the baseline 2 × 500 English → German model is 8.2 while the perplexity of the corresponding Seq-KD model is 22.7”
§ Sequence-Level Knowledge Distillation > 7 Conclusion
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We use beam size K = 5 (we did not see improvements with a larger beam).”
§ Sequence-Level Knowledge Distillation > Sequence-Level Knowledge Distillation (Seq-KD)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
In the paper
“Our student models can even be run efficiently on a standard smartphone.”
§ Sequence-Level Knowledge Distillation > 1 Introduction
What the sources say
- Not matchedcites 1 https://github.com/harvardnlp/nmt-android
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
- Not matched
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We find that we can run the student model 10 times faster with greedy decoding than the teacher model with beam search on GPU (1051.3vs 101.9 words/sec), with similar performance.”
§ Sequence-Level Knowledge Distillation > 5.1 Decoding Speed
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Seq-KD model does significantly better for both greedy (+4.2 BLEU) and beam search (+1.4 BLEU) decoding.”
§ Sequence-Level Knowledge Distillation > 7 Conclusion
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“While pruning 90% of the weights results in a more appreciable decrease of 1.0 BLEU, the model is drastically smaller with 8m parameters, which is 26× fewer than the original teacher model.”
§ Sequence-Level Knowledge Distillation > 5.2 Weight Pruning
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“indicating that these methods provide orthogonal means of transferring knowledge from the teacher to the student: Word-KD is transferring knowledge at the the local (i.e.word) level while Seq-KD is transferring knowledge at the global (i.e.sequence) level.”
§ Sequence-Level Knowledge Distillation > 5 Results and Discussion
hedgeddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
In the paper
“Weight pruning for NMT was recently investigated by See et al. (2016), who found that up to 80 -90% of the parameters in a large NMT model can be pruned with little loss in performance.”
§ Sequence-Level Knowledge Distillation > 5.2 Weight Pruning
What the sources say
- Not matchedcites See et al. (2016)
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
- Not matched
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Therefore, the fact that sequencelevel knowledge distillation allows for greedy decoding is significant, with practical implications for running NMT systems across various devices.”
§ Sequence-Level Knowledge Distillation > 5.1 Decoding Speed
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Rests on an assumption: Run-time complexity for beam search grows linearly with beam size.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The above methods are complementary and can be combined with each other.”
§ Sequence-Level Knowledge Distillation > Sequence-Level Interpolation (Seq-Inter)
plaincorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The size of word embeddings have little impact on run-time as the word embedding layer is a simple lookup table that only affects the first layer of the model.”
§ Sequence-Level Knowledge Distillation > 5.2 Weight Pruning
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: unclear — It is unclear why the paper says it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Despite its intractability, we posit that this sequence-level objective is worthwhile.It gives the teacher the chance to assign probabilities to complete sequences and therefore transfer a broader range of knowledge.”
§ Sequence-Level Knowledge Distillation > 3.2 Sequence-Level Knowledge Distillation
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence.
The authors state this as an assumption.
The teacher has previously been trained, and we are estimating parameters for the student.
“We generally assume that the teacher has previously been trained, and that we are estimating parameters for the student.”
The authors state this as an assumption.
The teacher distribution parameters remain fixed during student training.
“where θ T parameterizes the teacher distribution and remains fixed.”
The paper relies on this without support.
Run-time complexity for beam search grows linearly with beam size.
“Run-time complexity for beam search grows linearly with beam size.”
The paper relies on this without support.
Combining pruning and knowledge distillation has not been investigated before.
“To our knowledge combining pruning and knowledge distillation has not been investigated before.”
The paper relies on this without support.
The translation domain has generally required the largest capacity deep learning models.
“We have chosen to focus on translation as this domain has generally required the largest capacity deep learning models”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Extracting reasoning matters not because output-only distillation is ineffective, but because reasoning yields a substantially more effective form of capability stealing.”
From Stealing Reasoning Traces from Proprietary LLM APIs (2026) · cites this paper as Kim and Rush, 2016
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
“on the 2 × 500 English → German model the word embeddings account for approximately 63% (50m out of 84m) of the parameters.”
§ Sequence-Level Knowledge Distillation > 5.2 Weight Pruning
Reported 63 · recomputed 59.52381
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
“Student Baseline 2 × 500 (Params: 84m) Word-KD Seq-KD Baseline + Seq-Inter Word-KD + Seq-Inter Seq-KD + Seq-Inter Seq-KD + Word-KD Seq-KD + Seq-Inter + Word-KD | 14.7 15.4 18.9 18.5 18.3 18.9 18.7 18.8 | -+0.7 +4.2 +3.6”
§ Sequence-Level Knowledge Distillation > 7 Conclusion
Reported 3.6 · recomputed 3.8
Citations
Cites
Hochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318
Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation
Cho, K., van Merrienboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., & Bengio, Y. (2014) · Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.3115/v1/d14-1179 · Cited by 25,545
BLEU: a method for automatic evaluation of machine translation
Papineni, K., Roukos, S., Ward, T., & Zhu, W.-J. (2001) · Proceedings of the 40th Annual Meeting on Association for Computational Linguistics - ACL '02 · doi:10.3115/1073083.1073135 · Cited by 21,344
Sequence to Sequence Learning with Neural Networks
Sutskever, I., Vinyals, O., & Le, Q. V. (2014) · arXiv · doi:10.48550/arxiv.1409.3215 · Cited by 17,766
Neural Machine Translation by Jointly Learning to Align and Translate
Bahdanau, D., Cho, K., & Bengio, Y. (2014) · arXiv · doi:10.48550/arxiv.1409.0473 · Cited by 16,213
Distilling the Knowledge in a Neural Network
Hinton, G., Vinyals, O., & Dean, J. (2015) · arXiv · doi:10.48550/arxiv.1503.02531 · Cited by 14,204
Effective Approaches to Attention-based Neural Machine Translation
Luong, T., Pham, H., & Manning, C. D. (2015) · Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/d15-1166 · Cited by 8,382
Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
Xu, K., Ba, J., Kiros, R., Cho, K., Courville, A., Salakhutdinov, R., Zemel, R., & Bengio, Y. (2015) · arXiv · doi:10.48550/arxiv.1502.03044 · Cited by 8,248
Show and tell: A neural image caption generator
Vinyals, O., Toshev, A., Bengio, S., & Erhan, D. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298935 · Cited by 6,318
Han, S., Mao, H., & Dally, W. J. (2015) · arXiv · doi:10.48550/arxiv.1510.00149 · Cited by 4,662
Research works printed but not linked: 4
- Cristian Bucila, Rich Caruana, and Alexandru Niculescu-Mizil. 2006. Model Compres- sion. In Proceedings of KDD.
- and Blunsom2013] Nal Kalchbrenner and Phil Blunsom. 2013. Recurrent Continuous Transla- tion Models. In Proceedings of EMNLP.
- Jiwei Li, Michael Galley, Chris Brockett, Jianfeg Gao, and Bill Dolan. 2016. A Diversity- Promoting Objective Function for Neural Conversa- tional Models. In Proceedings of NAACL 2016.
- Mariet and Sra2016] Zelda Mariet and Suvrit Sra. 2016. Diversity Networks. In Proceedings of ICLR.
Cited by
Sorted by citations
Items 501–600 of 841
Tuna: Instruction Tuning using Feedback from Large Language Models
Li, H., Liu, Y., Zhang, X., Lu, W., & Wei, F. (2023) · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) · doi:10.18653/v1/2023.findings-emnlp.1011 · Cited by 14
An Overview of Uncertainty Calibration for Text Classification and the Role of Distillation
Guo, H., Pasunuru, R., & Bansal, M. (2021) · Workshop on Representation Learning for NLP (RepL4NLP) · doi:10.18653/v1/2021.repl4nlp-1.29 · Cited by 5
Dual knowledge distillation for bidirectional neural machine translation
Zhang, H., Qiu, S., & Wu, S. (2021) · doi:10.1109/ijcnn52387.2021.9534396 · Cited by 5
Building Multilingual Machine Translation Systems That Serve Arbitrary XY Translations
Eriguchi, A., Xie, S., Qin, T., et al. (2022) · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT) · doi:10.18653/v1/2022.naacl-main.44 · Cited by 5
HEFT: A History-Enhanced Feature Transfer framework for incremental event detection
Wei, K., Zhang, Z., Jin, L., Guo, Z. M., Li, S., Wang, W., & Lv, J. (2022) · Knowledge-Based Systems · doi:10.1016/j.knosys.2022.109601 · Cited by 6
Direct Speech-to-Text Translation Models as Students of Text-to-Text Models
Gaido, M., Negri, M., & Turchi, M. (2022) · Italian Journal of Computational Linguistics · doi:10.4000/ijcol.959 · Cited by 7
Multi-stage Progressive Compression of Conformer Transducer for On-device Speech Recognition
Rathod, J., Dawalatabad, N., Singh, S., & Gowda, D. (2022) · Interspeech (USB) · doi:10.21437/interspeech.2022-10582 · Cited by 5
Distilling Knowledge with a Teacher’s Multitask Model for Biomedical Named Entity Recognition
Mehmood, T., Gerevini, A., Lavelli, A., Olivato, M., & Serina, I. (2023) · Information · doi:10.3390/info14050255 · Cited by 6
Towards Real-Time Mispronunciation Detection in Kids' Speech
Plantinga, P., & Fosler‐Lussier, E. (2019) · IEEE Automatic Speech Recognition and Understanding Workshop (ASRU) · doi:10.1109/asru46091.2019.9003863 · Cited by 5
Knowledge Distillation Techniques for Biomedical Named Entity Recognition.
Mehmood, T., Serina, I., Lavelli, A., & Gerevini, A. (2020) · Cited by 5
The NiuTrans End-to-End Speech Translation System for IWSLT 2021 Offline Task
Xu, C., Liu, X., Liu, X., et al. (2021) · International Conference on Spoken Language Translation (IWSLT) · doi:10.18653/v1/2021.iwslt-1.9 · Cited by 5
Modeling Coverage for Non-Autoregressive Neural Machine Translation
Shan, Y., Feng, Y., & Shao, C. (2021) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn52387.2021.9533529 · Cited by 5
Torunoğlu-Selamet, D., İnceoğlu, A., & Eryiğit, G. (2020) · International Conference on Computer Science and Engineering (UBMK) · doi:10.1109/ubmk50275.2020.9219389 · Cited by 5
Gao, Y., Parcollet, T., & Lane, N. D. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2005.09310 · Cited by 5
Improved Mask-CTC for Non-Autoregressive End-to-End ASRPreprint
Higuchi, Y., Inaguma, H., Watanabe, S., Ogawa, T., & Kobayashi, T. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2010.13270 · Cited by 5
Multi-scale Joint Learning with Negative Sample Mining for Non-autoregressive Machine Translation
Qu, X., Liu, G., & Li, L. (2025) · Knowledge-Based Systems · doi:10.1016/j.knosys.2025.113610 · Cited by 5
Active Data Curation Effectively Distills Large-Scale Multimodal Models
Udandarao, V., Parthasarathy, N., Naeem, M. F., Evans, T., Albanie, S., Tombari, F., Xian, Y., Tonioni, A., & Hénaff, O. J. (2025) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52734.2025.01345 · Cited by 15
LLM-enhanced multi-level knowledge distillation for molecular property prediction
Zhuang, L., Sun, Y., Jun, Z., Lu, Y., & Wang, H. (2025) · Knowledge-Based Systems · doi:10.1016/j.knosys.2025.114195 · Cited by 4
Theoretical Perspectives on Knowledge Distillation: A Review
Liu, C., Yin, H., & Wang, X. (2025) · Wiley Interdisciplinary Reviews Computational Statistics · doi:10.1002/wics.70049 · Cited by 6
m2LLM: A Multi-Dimensional Optimization Framework for LLM Inference on Mobile Devices
Liu, K., Zhou, X., & Li, L. (2025) · IEEE Transactions on Parallel and Distributed Systems · doi:10.1109/tpds.2025.3587445 · Cited by 7
Rephrasing the Reference for Non-autoregressive Machine Translation
Shao, C., Zhang, J., Zhou, J., & Feng, Y. (2023) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v37i11.26587 · Cited by 5
Select, Prompt, Filter: Distilling Large Language Models for Summarizing Conversations
Pham, M.-Q., Indurthi, S. R., Chollampatt, S., & Turchi, M. (2023) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2023.emnlp-main.753 · Cited by 5
AIST AIRC Submissions to the WMT23 Shared Task
Rikters, M., & Miwa, M. (2023) · Conference on Machine Translation (WMT) · doi:10.18653/v1/2023.wmt-1.13 · Cited by 4
Ding, H., Chen, K., & Huo, Q. (2019) · Interspeech (USB) · doi:10.21437/interspeech.2019-2182 · Cited by 4
Tencent Minority-Mandarin Translation System
Hu, B., Han, A., Zhang, Z., Huang, S., & Ju, Q. (2019) · Communications in computer and information science · doi:10.1007/978-981-15-1721-1_10 · Cited by 4
Khayrallah, H., Bremerman, J., McCarthy, A. D., Murray, K., Wu, W. M. C., & Post, M. (2020) · Workshop on Neural Generation and Translation (WNGT) · doi:10.18653/v1/2020.ngt-1.22 · Cited by 4
Antaris, S., Rafailidis, D., & Girdzijauskas, Š. (2020) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata50022.2020.9378219 · Cited by 4
Improving Stylized Neural Machine Translation with Iterative Dual Knowledge Transfer
Wu, X., Liu, J., Li, X., Xu, J., Chen, Y., Zhang, Y., & Huang, H. (2021) · International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2021/547 · Cited by 4
Kim, B., Seo, S., Han, S., Erdenee, E., & Chang, B. (2021) · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) · doi:10.18653/v1/2021.findings-emnlp.286 · Cited by 6
Knowledge distillation on neural networks for evolving graphs
Antaris, S., Rafailidis, D., & Girdzijauskas, Š. (2021) · Social Network Analysis and Mining · doi:10.1007/s13278-021-00816-1 · Cited by 5
Model Compression Techniques in Deep Neural Networks
Yesuf, M. M., & Assefa, B. G. (2023) · Communications in computer and information science · doi:10.1007/978-3-031-31327-1_10 · Cited by 4
Zhou, W., Bras, R. J. L., & Choi, Y. (2023) · Findings of the Association for Computational Linguistics: ACL (Findings of ACL) · doi:10.18653/v1/2023.findings-acl.664 · Cited by 5
Revisiting Non-Autoregressive Translation at Scale
Wang, Z., Wang, L., Su, J., Yao, J., & Tu, Z. (2023) · Findings of the Association for Computational Linguistics: ACL (Findings of ACL) · doi:10.18653/v1/2023.findings-acl.763 · Cited by 4
DePA: Improving Non-autoregressive Translation with Dependency-Aware Decoder
Zhan, J., Chen, Q., Chen, B., Wang, W., Bai, Y., & Gao, Y. (2023) · International Conference on Spoken Language Translation (IWSLT) · doi:10.18653/v1/2023.iwslt-1.47 · Cited by 9
Reducing Model Churn: Stable Re-training of Conversational Agents
Hidey, C., Liu, F., & Goel, R. (2022) · Annual Meeting of the Special Interest Group on Discourse and Dialogue · doi:10.18653/v1/2022.sigdial-1.2 · Cited by 4
Non-autoregressive Streaming Transformer for Simultaneous Translation
Ma, Z., Zhang, S., Guo, S., Shao, C., Zhang, M., & Feng, Y. (2023) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2023.emnlp-main.314 · Cited by 8
Self-Improvement of Non-autoregressive Model via Sequence-Level Distillation
Liao, Y., Jiang, S., Li, Y., Wang, Y., & Wang, Y. (2023) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2023.emnlp-main.878 · Cited by 4
Rule-based Morphological Inflection Improves Neural Terminology Translation
Xu, W., & Carpuat, M. J. (2021) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2021.emnlp-main.477 · Cited by 4
Semi-supervised Music Tagging Transformer
Won, M., Choi, K., & Serra, X. (2021) · Repositori digital de la UPF (Universitat Pompeu Fabra) · doi:10.5281/zenodo.5624405 · Cited by 4
The AISP-SJTU Simultaneous Translation System for IWSLT 2022
Zhu, Q., Wu, R., Liu, G., Zhu, X., Chen, X., Zhou, Y., Miao, Q., Wang, R., & Yu, K. (2022) · International Conference on Spoken Language Translation (IWSLT) · doi:10.18653/v1/2022.iwslt-1.16 · Cited by 4
The Xiaomi Text-to-Text Simultaneous Speech Translation System for IWSLT 2022
Guo, B., Liu, M., Zhang, W., Chen, H., Mu, C., Li, X., Cui, J., Wang, B., & Guo, Y. (2022) · International Conference on Spoken Language Translation (IWSLT) · doi:10.18653/v1/2022.iwslt-1.17 · Cited by 4
Amortized Noisy Channel Neural Machine Translation
Pang, R. Y., He, H., & Cho, K. (2022) · International Conference on Natural Language Generation · doi:10.18653/v1/2022.inlg-main.11 · Cited by 4
Balancing Cost and Benefit with Tied-Multi Transformers
Dabre, R., Rubino, R., & Fujita, A. (2020) · Workshop on Neural Generation and Translation (WNGT) · doi:10.18653/v1/2020.ngt-1.3 · Cited by 7
Multilingual Speech Translation KIT @ IWSLT2021
Pham, N.-Q., Nguyen, T. N., Ha, T.-L., et al. (2021) · International Conference on Spoken Language Translation (IWSLT) · doi:10.18653/v1/2021.iwslt-1.18 · Cited by 4
Non-autoregressive Neural Machine Translation with Distortion Model
Zhou, L., Zhang, J., Zhao, Y., & Zong, C. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-60450-9_32 · Cited by 4
Mun'im, R. M., Inoue, N., & Shinoda, K. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1811.04531 · Cited by 4
Knowledge Distillation in Document RetrievalPreprint
Shakeri, S., Sethy, A., & Cheng, C. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1911.11065 · Cited by 6
Teaching Machines to ConversePreprint
Li, J. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2001.11701 · Cited by 5
Power Consumption Variation over Activation FunctionsPreprint
Derczynski, L. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.07237 · Cited by 10
Faster Re-translation Using Non-Autoregressive Model For Simultaneous Neural Machine TranslationPreprint
Han, H., Indurthi, S. R., Zaidi, M. A., Lakumarapu, N. K., Lee, B., Kim, S., Kim, C., & Hwang, I.-C. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2012.14681 · Cited by 5
Non-autoregressive End-to-end Speech Translation with Parallel Autoregressive RescoringPreprint
Inaguma, H., Higuchi, Y., Duh, K., Kawahara, T., & Watanabe, S. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2109.04411 · Cited by 10
Research on Multilingual News Clustering Based on Cross-Language Word EmbeddingsPreprint
Wu, L., Li, R., & Lam, W.-H. (2023) · Preprints.org · doi:10.20944/preprints202305.0990.v1 · Cited by 4
Can Advanced LLMs Coach Smaller LLMs? Knowledge Distillation for Goal-Oriented DialogsPreprint
Wang, T., Sudhir, K., & Hong, D. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4924881 · Cited by 4
BitextEdit: Automatic Bitext Editing for Improved Low-Resource Machine Translation
Briakou, E., Wang, S. I., Zettlemoyer, L., & Ghazvininejad, M. (2022) · Findings of the Association for Computational Linguistics: NAACL · doi:10.18653/v1/2022.findings-naacl.110 · Cited by 3
N-Gram Nearest Neighbor Machine Translation
Lv, R., Guo, J., Wang, R., Tan, X., Qin, T., & Liu, Q. (2024) · IEEE Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2024.3407537 · Cited by 4
Scaling Motion Forecasting Models with Ensemble Distillation
Ettinger, S. M., Goel, K., Srivastava, A., et al. (2024) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra57147.2024.10611196 · Cited by 3
Learning to Count Without Annotations
Knobel, L., Han, T., & Asano, Y. M. (2024) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52733.2024.02163 · Cited by 5
Knowledge Distillation Using Soft and Hard Labels and Annealing for Acoustic Model Training
Tachioka, Y. (2019) · IEEE Global Conference on Consumer Electronics (GCCE) · doi:10.1109/gcce46687.2019.9015500 · Cited by 3
Lightspeech: Lightweight Non-Autoregressive Multi-Speaker Text-To-Speech
Li, S., Ouyang, B., Li, L., et al. (2021) · IEEE Spoken Language Technology Workshop (SLT) · doi:10.1109/slt48900.2021.9383562 · Cited by 3
Veri Artırımı için Yarı-Denetimli Bağlamsal Anlam Belirsizliği Giderme
Torunoğlu-Selamet, D., & Eryiğit, G. (2021) · Türkiye Bilişim Vakfı Bilgisayar Bilimleri ve Mühendisliği Dergisi · doi:10.54525/tbbmd.835744 · Cited by 3
Target-Oriented Knowledge Distillation with Language-Family-Based Grouping for Multilingual NMT
Do, H., & Lee, G. G. (2022) · ACM Transactions on Asian and Low-Resource Language Information Processing · doi:10.1145/3546067 · Cited by 3
Integrating Translation Memories into Non-Autoregressive Machine Translation
Xu, J., Crego, J., & Yvon, F. (2023) · Conference of the European Chapter of the Association for Computational Linguistics · doi:10.18653/v1/2023.eacl-main.96 · Cited by 3
Mutually Distilled Sparse RCNN for Few-Shot Object Detection
Jiang, X., Yu, H.-C., Lv, Y., & Zhu, X. F. (2022) · IEEE International Conference on Frontiers Technology of Information and Computer (ICFTIC) · doi:10.1109/icftic57696.2022.10075197 · Cited by 2
Towards Domain Generalisation in ASR with Elitist Sampling and Ensemble Knowledge Distillation
Ahmad, R., Jalal, M. A., Farooq, M. U., Ollerenshaw, A., & Hain, T. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10095746 · Cited by 3
When Gradient Descent Meets Derivative-Free Optimization: A Match Made in Black-Box Scenario
Han, C., Cui, L., Zhu, R., Wang, J., Chen, N., Sun, Q., Li, X., & Gao, M. (2023) · Findings of the Association for Computational Linguistics: ACL (Findings of ACL) · doi:10.18653/v1/2023.findings-acl.55 · Cited by 5
Commonsense Knowledge Transfer for Pre-trained Language Models
Zhou, W., Bras, R. J. L., & Choi, Y. (2023) · Findings of the Association for Computational Linguistics: ACL (Findings of ACL) · doi:10.18653/v1/2023.findings-acl.368 · Cited by 6
The USTC’s Dialect Speech Translation System for IWSLT 2023
Deng, P., Chen, S., Zhang, W., et al. (2023) · International Conference on Spoken Language Translation (IWSLT) · doi:10.18653/v1/2023.iwslt-1.5 · Cited by 3
Self-training Reduces Flicker in Retranslation-based Simultaneous Translation
Sen, S., Sennrich, R., Zhang, B., & Haddow, B. (2023) · Conference of the European Chapter of the Association for Computational Linguistics · doi:10.18653/v1/2023.eacl-main.270 · Cited by 3
Al-Dahlaki, N. O. M., & Al-Qaysi, S. A. S. (2023) · Journal of the Faculty of Medicine Baghdad · doi:10.32007/jfacmedbagdad.2078 · Cited by 3
Exploring iterative dual domain adaptation for neural machine translation
Liu, X., Zeng, J., Wang, X., Wang, Z., & Su, J. (2023) · Knowledge-Based Systems · doi:10.1016/j.knosys.2023.111182 · Cited by 3
Enhancing Abstractiveness of Summarization Models through Calibrated Distillation
Song, H., Shalyminov, I., Su, H., Singh, S., Yao, K., & Mansour, S. (2023) · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) · doi:10.18653/v1/2023.findings-emnlp.468 · Cited by 4
Towards Example-Based NMT with Multi-Levenshtein Transformers
Bouthors, M., Crego, J., & Yvon, F. (2023) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2023.emnlp-main.113 · Cited by 3
Multilingual Non-Autoregressive Machine Translation without Knowledge Distillation
Huang, C., Huang, F., Zheng, Z., Zai͏̈ane, O. R., Zhou, H., & Mou, L. (2023) · doi:10.18653/v1/2023.findings-ijcnlp.14 · Cited by 5
Leveraging Diverse Modeling Contexts With Collaborating Learning for Neural Machine Translation
Liao, Y., Wang, Y., & Wang, Y. (2024) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2024.3379884 · Cited by 3
Parameter Efficient Diverse Paraphrase Generation Using Sequence-Level Knowledge Distillation
Jayawardena, L., & Yapa, P. (2024) · International Conference on Advancements in Computational Sciences (ICACS) · doi:10.1109/icacs60934.2024.10473289 · Cited by 3
Applied Federated Model Personalization in the Industrial Domain: A Comparative Study
Siniosoglou, I., Argyriou, V., Fragulis, G. F., Fouliras, P., Papadopoulos, G. T., Lytos, A., & Sarigiannidis, P. G. (2024) · IEEE Open Journal of the Communications Society · doi:10.1109/ojcoms.2024.3457803 · Cited by 3
Boosting Diary Study Outcomes with a Fine-Tuned Large Language Model
Oh, S.-Y., Zhao, J., Russo, C., & Bolmer, M. (2025) · CHI Conference Extended Abstracts on Human Factors in Computing Systems (CHI EA) · doi:10.1145/3706599.3719287 · Cited by 3
A Regularization-based Framework for Bilingual Grammar Induction
Jiang, Y., Han, W., & Tu, K. (2019) · doi:10.18653/v1/d19-1148 · Cited by 3
Deep Transfer Learning for Search and Recommendation
Li, Y., Yang, Y., Zhou, S., et al. (2020) · The Web Conference (WWW) · doi:10.1145/3366424.3383115 · Cited by 3
SMRT Chatbots: Improving Non-Task-Oriented Dialog with Simulated Multiple Reference Training
Khayrallah, H., & Sedoc, J. (2020) · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) · doi:10.18653/v1/2020.findings-emnlp.403 · Cited by 3
Leveraging Unlabeled Speech for Sequence Discriminative Training of Acoustic Models
Sapru, A., & Garimella, S. (2020) · Interspeech (USB) · doi:10.21437/interspeech.2020-2056 · Cited by 3
Unifying Input and Output Smoothing in Neural Machine Translation
Gao, Y., Liao, B., & Ney, H. (2020) · International Conference on Computational Linguistics · doi:10.18653/v1/2020.coling-main.386 · Cited by 3
Teacher and Student Models of Offensive Language in Social Media
Ranasinghe, T., & Zampieri, M. (2023) · Findings of the Association for Computational Linguistics: ACL (Findings of ACL) · doi:10.18653/v1/2023.findings-acl.241 · Cited by 3
Recursive Top-Down Production for Sentence Generation with Latent Trees
Tan, S., Shen, Y., Sordoni, A., Courville, A. C., & O’Donnell, T. J. (2020) · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) · doi:10.18653/v1/2020.findings-emnlp.208 · Cited by 3
Edinburgh’s End-to-End Multilingual Speech Translation System for IWSLT 2021
Zhang, B., & Sennrich, R. (2021) · International Conference on Spoken Language Translation (IWSLT) · doi:10.18653/v1/2021.iwslt-1.19 · Cited by 3
On Machine Translation of User Reviews
Popović, M., Way, A., Poncelas, A., Rakuten Asia, S., & Bakarić, M. B. (2021) · International conference Recent advances in natural language processing · doi:10.26615/978-954-452-072-4_124 · Cited by 4
Analysis of Knowledge Transfer in Kernel Regime
Panahi, A., Rahbar, A., Bhattacharyya, C., Dubhashi, D., & Chehreghani, M. H. (2022) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3511808.3557237 · Cited by 3
Submission of USTC’s System for the IWSLT 2023 - Offline Speech Translation Track
Zhou, X., Cui, J., Ye, Z., et al. (2023) · International Conference on Spoken Language Translation (IWSLT) · doi:10.18653/v1/2023.iwslt-1.15 · Cited by 3
Layer-Wise Multi-View Learning for Neural Machine Translation
Wang, Q., Li, C., Zhang, Y., et al. (2020) · International Conference on Computational Linguistics · doi:10.18653/v1/2020.coling-main.377 · Cited by 3
Fukuda, R., Oka, Y., Kano, Y., Yano, Y., Ko, Y., Tokuyama, H., Doi, K., Sakti, S., Sudoh, K., & Nakamura, S. (2021) · International Conference on Spoken Language Translation (IWSLT) · doi:10.18653/v1/2021.iwslt-1.3 · Cited by 3
Self-Training Sampling with Monolingual Data Uncertainty for Neural Machine TranslationPreprint
Jiao, W., Wang, X., Tu, Z., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.00941 · Cited by 3
de Gibert, O., Vázquez, R., Aulamo, M., Scherrer, Y., Virpioja, S., & Tiedemann, J. (2023) · Workshop on Natural Language Processing for Indigenous Languages of the Americas (AmericasNLP) · doi:10.18653/v1/2023.americasnlp-1.20 · Cited by 2
Liu, Y., Yu, H., Li, H., Han, X., Zhang, X., & Liu, H. (2025) · IEEE Internet of Things Journal · doi:10.1109/jiot.2025.3579587 · Cited by 2
Nath, B., & Gulzar, Y. (2026) · Scientific Reports · doi:10.1038/s41598-025-33794-x · Cited by 2
Obtain Dark Knowledge via Extended Knowledge Distillation
Chen, Y., & Pan, R. (2019) · International Conference on Artificial Intelligence and Advanced Manufacturing (AIAM) · doi:10.1109/aiam48774.2019.00106 · Cited by 2
Hierarchical Knowledge Distillation for Dialogue Sequence Labeling
Orihashi, S., Yamazaki, Y., Makishima, N., Ihori, M., Takashima, A., Tanaka, T., & Masumura, R. (2021) · IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU) · doi:10.1109/asru51503.2021.9687959 · Cited by 2
Metadata
- Authors
- Kim, Y., & Rush, A. M.
- Affiliations
- Harvard University (US)
- Venue
- Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing
- Published
- 2016
- Pages
- 1317-1327
- OpenAlex
- W2963736842
- Semantic Scholar
- 57a10537978600fd33dcdd48922c791609a4851a
- Cited by
- 836
- References
- 53
- Citations by year
- 2026: 36 · 2025: 59 · 2024: 52 · 2023: 137 · 2022: 115 · 2021: 157 · 2020: 143 · 2019: 75 · 2018: 42 · 2017: 16 · 2016: 4
- FWCI
- 24.59
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- Locations
- Conference on Empirical Methods in Natural Language Processing (EMNLP) (publishedVersion, cc-by)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Keywords
- knowledge distillation, neural machine translation, model compression, weight pruning, teacher-student models, parameter reduction
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-23, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-23 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
