mlsanityData to 7 Oct 2026

As Little as Possible, as Much as Necessary: Detecting Over- and Undertranslations with Contrastive Conditioning

arXiv 2203.01927 ↗First posted 3 Mar 2022cs.CLcited by 11

Not on the map. Why?

Abstract (arXiv)

Omission and addition of content is a typical issue in neural machine translation. We propose a method for detecting such phenomena with off-the-shelf translation models. Using contrastive conditioning, we compare the likelihood of a full sequence under a translation model to the likelihood of its parts, given the corresponding source or target sequence. This allows to pinpoint superfluous words in the translation and untranslated words in the source even in the absence of a reference translation. The accuracy of our method is comparable to a supervised method that requires a custom quality estimation model.

Key facts. Cited by 11 papers in the map (data through 7 Oct 2026). Builds on 13 of its 33 references that are in the map. First posted on arXiv 3 Mar 2022.

Builds on

13 of 33 references.

References · 13

  1. 2019BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehensioncited by 5,874
  2. 2019Unsupervised Cross-lingual Representation Learning at Scalecited by 3,907
  3. 2016Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translationcited by 1,471
  4. 2020Multilingual Denoising Pre-training for Neural Machine Translationcited by 1,028
  5. 2020Stanza: A Python Natural Language Processing Toolkit for Many Human Languagescited by 719
  6. 2021Experts, Errors, and Context: A Large-Scale Study of Human Evaluation for Machine Translationcited by 333
  7. 2020Detecting Hallucinated Content in Conditional Neural Sequence Generationcited by 151
  8. 2016Modeling Coverage for Neural Machine Translationcited by 110
  9. 2019OpenKiwi: An Open Source Framework for Quality Estimationcited by 49
  10. 2021Quality Estimation without Human-labeled Datacited by 15
  11. 2018Neural Machine Translation with Adequacy-Oriented Learningcited by 8
  12. 2017Modeling Past and Future for Neural Machine Translationcited by 6
1 more
  1. 2018Otem&Utem: Over- and Under-Translation Evaluation Metric for NMTcited by 2

Built on by

Cited by year

Citing papers per year: 2022: 1; 2023: 5; 2024: 3; 2025: 2.

11 papers.

  1. 2023Mitigating Hallucinations and Off-target Machine Translation with Source-Contrastive and Language-Contrastive Decodingcited by 42
  2. 2023HalOmi: A Manually Annotated Benchmark for Multilingual Hallucination and Omission Detection in Machine Translationcited by 35
  3. 2024Multi-property Steering of Large Language Models with Dynamic Activation Compositioncited by 35
  4. 2022ACES: Translation Accuracy Challenge Sets for Evaluating Machine Translation Metricscited by 25
  5. 2024Word Alignment as Preference for Machine Translationcited by 14
  6. 2024MMTE: Corpus and Metrics for Evaluating Machine Translation Quality of Metaphorical Languagecited by 10
  7. 2025Are We Paying Attention to Her? Investigating Gender Disambiguation and Attention in Machine Translationcited by 7
  8. 2023Quantifying the Plausibility of Context Reliance in Neural Machine Translationcited by 6
  9. 2023Towards Fine-Grained Information: Identifying the Type and Location of Translation Errorscited by 4
  10. 2025M2PO: Multi-Perspective Multi-Pair Preference Optimization for Machine Translationcited by 3
  11. 2023With a Little Help from the Authors: Reproducing Human Evaluation of an MT Error Detectorcited by 0