Differentiable Scheduled Sampling for Credit Assignment
Open accessAbstract
We demonstrate that a continuous relaxation of the argmax operation can be used to create a differentiable approximation to greedy decoding for sequence-tosequence (seq2seq) models.By incorporating this approximation into the scheduled sampling training procedure (Bengio et al., 2015)-a well-known technique for correcting exposure bias-we introduce a new training objective that is continuous and differentiable everywhere and that can provide informative gradients near points where previous decoding decisions change their value.In addition, by using a related approximation, we demonstrate a similar approach to sampled-based training.Finally, we show that our approach outperforms cross-entropy training and scheduled sampling procedures in two sequence prediction tasks: named entity recognition and machine translation.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
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 PersonaChat datasets have accelerated research into persona consistency.
“The PersonaChat datasets [47, 312] have accelerated research into persona consistency [96, 128, 178, 284, 296, 307, 318].”
From Survey of Hallucination in Natural Language Generation (2023) · cites this paper as [47]
The Experimental Setup (Section 5; page unavailable) describes machine-translation and named-entity-recognition datasets, not PersonaChat. Searches for PersonaChat and persona consistency found no passage addressing whether those datasets accelerated research. (LLM, conf=0.78)
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
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 · see citing works
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 · see citing works
Categorical Reparameterization with Gumbel-Softmax
Jang, E., Gu, S., & Poole, B. (2016) · arXiv · doi:10.48550/arxiv.1611.01144 · Cited by 3,616 · see citing works
A Neural Attention Model for Abstractive Sentence Summarization
Rush, A. M., Chopra, S., & Weston, J. (2015) · Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/d15-1044 · Cited by 2,879 · see citing works
A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning
Ross, S., Gordon, G. J., & Bagnell, J. A. (2018) · Carnegie Mellon University · doi:10.1184/r1/6550949 · Cited by 1,941 · see citing works
Building End-To-End Dialogue Systems Using Generative Hierarchical Neural Network Models
Serban, I. V., Sordoni, A., Bengio, Y., Courville, A. C., & Pineau, J. (2016) · AAAI Publications (The Association for the Advancement of Artificial Intelligence (AAAI)) · doi:10.1609/aaai.v30i1.9883 · Cited by 1,639 · see citing works
Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks
Bengio, S., Vinyals, O., Jaitly, N., & Shazeer, N. (2015) · arXiv · doi:10.48550/arxiv.1506.03099 · Cited by 1,621 · see citing works
Sequence Level Training with Recurrent Neural Networks
Ranzato, M., Chopra, S., Auli, M., & Zaremba, W. (2016) · International Conference on Learning Representations · Cited by 572 · see citing works
Globally Normalized Transition-Based Neural Networks
Andor, D., Alberti, C., Weiss, D., Severyn, A., Presta, A., Ganchev, K., Petrov, S., & Collins, M. (2016) · Proceedings of the conference - Association for Computational Linguistics. Meeting · doi:10.18653/v1/p16-1231 · Cited by 494 · see citing works
Sequence-to-Sequence Learning as Beam-Search Optimization
Wiseman, S., & Rush, A. M. (2016) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/d16-1137 · Cited by 485 · see citing works
An Actor-Critic Algorithm for Sequence Prediction
Bahdanau, D., Brakel, P., Xu, K., Goyal, A., Lowe, R., Pineau, J., Courville, A., & Bengio, Y. (2016) · arXiv · doi:10.48550/arxiv.1607.07086 · Cited by 423 · see citing works
Learning to Transduce with Unbounded MemoryPreprint
Grefenstette, E., Hermann, K. M., Suleyman, M., & Blunsom, P. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1506.02516 · Cited by 221 · see citing works
Report on the 11th IWSLT Evaluation Campaign, IWSLT 2014
Cettolo, M., Niehues, J., Stüker, S., Bentivogli, L., & Federico, M. (2014) · Cited by 158 · see citing works
Learning as search optimization
Daumé, H., & Marcu, D. (2005) · International Conference on Machine Learning · doi:10.1145/1102351.1102373 · Cited by 157 · see citing works
Approximation-Aware Dependency Parsing by Belief Propagation
Gormley, M. R., Dredze, M. H., & Eisner, J. M. (2015) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00153 · Cited by 25 · see citing works
Decoding as Continuous Optimization in Neural Machine TranslationPreprint
Hoang, C. D. V., Haffari, G., & Cohn, T. (2017) · arXiv (Cornell University) · doi:10.13140/rg.2.2.36626.86720 · Cited by 7 · see citing works
Towards Decoding as Continuous Optimization in Neural Machine TranslationPreprint
Hoang, C. D. V., Haffari, G., & Cohn, T. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1701.02854 · Cited by 6 · see citing works
Flagged first, then by citations
1 of the 18 references couldn’t be matched to a known paper.
Research works printed but not linked: 3
- Hal Daumé, John Langford, and Daniel Marcu. 2009. Search-based structured prediction. Machine learn- ing 75(3):297-325.
- Chris J Maddison, Andriy Mnih, and Yee Whye Teh. 2017. The concrete distribution: A continuous re- laxation of discrete random variables. In Interna- tional Conference on Learning Representations.
- Erik F Tjong Kim Sang and Fien De Meulder. 2003. Introduction to the conll-2003 shared task: Language-independent named entity recognition. In Proceedings of the seventh conference on Natural language learning at HLT-NAACL 2003-Volume 4. Association for Computational Linguistics, pages 142-147.
Cited by
Sorted by citations
Survey of Hallucination in Natural Language Generation
Ji, Z., Lee, N., Frieske, R., Yu, T., Su, D., Xu, Y., Ishii, E., Bang, Y. J., Madotto, A., & Fung, P. (2023) · ACM Computing Surveys · doi:10.1145/3571730 · Cited by 5,025 · see citing works
“[47] create a diferentiable approximation to greedy decoding that shows a good performance in the NMT task.”
BackgroundStyle Transfer from Non-Parallel Text by Cross-AlignmentPreprint
Shen, T., Lei, T., Barzilay, R., & Jaakkola, T. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1705.09655 · Cited by 616 · see citing works
Deep Reinforcement Learning for Sequence-to-Sequence Models
Keneshloo, Y., Shi, T., Ramakrishnan, N., & Reddy, C. K. (2019) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2019.2929141 · Cited by 230 · see citing works
Context-aware Retrieval-based Deep Commit Message Generation
Wang, H., Xia, X., Lo, D., He, Q., Wang, X., & Grundy, J. (2021) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3464689 · Cited by 71 · see citing works
Scheduled Sampling for Transformers
Mihaylova, T., & Martins, A. F. T. (2019) · Annual Meeting of the Association for Computational Linguistics: Student Research Workshop (ACL SRW) · doi:10.18653/v1/p19-2049 · Cited by 69 · see citing works
Baziotis, C., Androutsopoulos, I., Konstas, I., & Potamianos, A. (2019) · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT) · doi:10.18653/v1/n19-1071 · Cited by 51 · see citing works
Towards Decoding as Continuous Optimisation in Neural Machine Translation
Hoang, C. D. V., Haffari, G., & Cohn, T. (2017) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/d17-1014 · Cited by 37 · see citing works
Baziotis, C., Androutsopoulos, I., Konstas, I., & Potamianos, A. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.03651 · Cited by 27 · see citing works
Autoregressive Knowledge Distillation through Imitation Learning
Lin, A., Wohlwend, J., Chen, H., & Lei, T. (2020) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2020.emnlp-main.494 · Cited by 24 · see citing works
Scheduled Sampling Based on Decoding Steps for Neural Machine Translation
Liu, Y., Meng, F., Chen, Y., Xu, J., & Zhou, J. (2021) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2021.emnlp-main.264 · Cited by 14 · see citing works
A Lightweight Framework for Fast Trajectory Simplification
Fang, Z., He, C., Chen, L., Hu, D., Sun, Q., Li, L., & Gao, Y. (2023) · IEEE International Conference on Data Engineering (ICDE) · doi:10.1109/icde55515.2023.00184 · Cited by 12 · see citing works
Confidence-Aware Scheduled Sampling for Neural Machine Translation
Liu, Y., Meng, F., Chen, Y., Xu, J., & Zhou, J. (2021) · Findings of the Association for Computational Linguistics: ACL (Findings of ACL) · doi:10.18653/v1/2021.findings-acl.205 · Cited by 12 · see citing works
Non-Autoregressive Neural Dialogue GenerationPreprint
Han, Q., Meng, Y., Wu, F. Q., & Li, J. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2002.04250 · Cited by 12 · see citing works
Evaluating Performance and Accuracy Improvements for Attention-OCR
Brzeski, A., Grinholc, K., Nowodworski, K., & Przybyłek, A. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-28957-7_1 · Cited by 11 · see citing works
Sparsifying Transformer Models with Trainable Representation Pooling
Pietruszka, M., Borchmann, Ł., & Garncarek, Ł. (2022) · Proceedings of the conference - Association for Computational Linguistics. Meeting · doi:10.18653/v1/2022.acl-long.590 · Cited by 10 · see citing works
Mixed Cross Entropy Loss for Neural Machine TranslationPreprint
Li, H., & Lu, W. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.15880 · Cited by 10 · see citing works
Differentiable Sampling with Flexible Reference Word Order for Neural Machine Translation
Xu, W., Niu, X., & Carpuat, M. J. (2019) · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT) · doi:10.18653/v1/n19-1207 · Cited by 9 · see citing works
Goyal, K., Dyer, C., & Berg-Kirkpatrick, T. (2019) · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT) · doi:10.18653/v1/n19-1171 · Cited by 9 · see citing works
Self-citationLAVA NAT: A Non-Autoregressive Translation Model with Look-Around Decoding and Vocabulary AttentionPreprint
Li, X., Meng, Y., Yuan, A., Wu, F., & Li, J. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2002.03084 · Cited by 9 · see citing works
A Continuous Relaxation of Beam Search for End-to-End Training of Neural Sequence Models
Goyal, K., Neubig, G., Dyer, C., & Berg-Kirkpatrick, T. (2018) · AAAI Publications (The Association for the Advancement of Artificial Intelligence (AAAI)) · doi:10.1609/aaai.v32i1.11806 · Cited by 7 · see citing works
Self-citationNoise Isn’t Always Negative: Countering Exposure Bias in Sequence-to-Sequence Inflection Models
Nicolai, G., & Silfverberg, M. (2020) · International Conference on Computational Linguistics · doi:10.18653/v1/2020.coling-main.255 · Cited by 7 · see citing works
CLIK-Diffusion: Clinical Knowledge-informed Diffusion Model for Tooth Alignment
Dou, Y., Wu, H. C., Li, C., Wang, C., Yang, T., Zhu, M., Shen, D. G., & Cui, Z. (2025) · Medical Image Analysis · doi:10.1016/j.media.2025.103746 · Cited by 6 · see citing works
Towards Decoding as Continuous Optimization in Neural Machine TranslationPreprint
Hoang, C. D. V., Haffari, G., & Cohn, T. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1701.02854 · Cited by 6 · see citing works
Language Model Evaluation in Open-ended Text GenerationPreprint
Nguyen, A. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2108.03578 · Cited by 5 · see citing works
An Integrated Interactive Framework for Natural Language to SQL Translation
Fan, Y., Ren, T., Guo, D., Zhao, Z., He, Z., Wang, X. S., Wang, Y., & Tao, S. (2023) · Lecture notes in computer science · doi:10.1007/978-981-99-7254-8_50 · Cited by 4 · see citing works
Masumura, R., Sato, H. P., Tanaka, T., Moriya, T., Ijima, Y., & Oba, T. (2019) · Interspeech (USB) · doi:10.21437/interspeech.2019-2111 · Cited by 3 · see citing works
Goyal, K., Dyer, C., & Berg-Kirkpatrick, T. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.06834 · Cited by 3 · see citing works
Self-citationMaking Classical Machine Learning Pipelines Differentiable: A Neural Translation ApproachPreprint
Yu, G.-I., Amizadeh, S., Kim, S., Pagnoni, A., Chun, B.-G., Weimer, M., & Interlandi, M. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1906.03822 · Cited by 2 · see citing works
Alleviate Exposure Bias in Sequence Prediction \\ with Recurrent Neural NetworksPreprint
Yuan, L., Feng, J., Zheng, X., & Huang, X. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2103.11603 · Cited by 2 · see citing works
Improving Scheduled Sampling with Elastic Weight Consolidation for Neural Machine Translation
Korakakis, M., & Vlachos, A. (2022) · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) · doi:10.18653/v1/2022.findings-emnlp.536 · Cited by 1 · see citing works
Reinforcement Replaces Supervision: Query focused Summarization using Deep Reinforcement Learning
Nath, S., Bhattacharyya, P., & Khadilkar, H. (2023) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2023.emnlp-main.977 · Cited by 1 · see citing works
Error-Correcting Neural Sequence PredictionPreprint
O’Neill, J., & Bollegala, D. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1901.07002 · Cited by 1 · see citing works
Differentiable Sampling with Flexible Reference Word Order for Neural Machine TranslationPreprint
Xu, W., Niu, X., & Carpuat, M. J. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.04079 · Cited by 1 · see citing works
Korakakis, M., & Vlachos, A. (2021) · arXiv (Cornell University)
Pictorial composition : modeling, perception & creation
Lelièvre, P. (2022) · theses.fr (ABES) · doi:10.70675/cc153babzed95z4048z9594z3b1b65403dbe
- Authors
- Goyal, K., Dyer, C., & Berg-Kirkpatrick, T.
- Affiliations
- Amazon (United States) (US)
- Carnegie Mellon University (US)
- Venue
- Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)
- Published
- 2017
- Pages
- 366-371
- ISSN
- 0736-587X (ISSN-L)
- OpenAlex
- W2607593044
- Cited by
- 34
- References
- 18
- Citations by year
- 2025: 1 · 2023: 3 · 2022: 4 · 2021: 5 · 2020: 4 · 2019: 12 · 2018: 2 · 2017: 3
- FWCI
- 2.62
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- Locations
- Proceedings of the conference - Association for Computational Linguistics. Meeting (publishedVersion, cc-by)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
- Physical Sciences › Computer Science › Artificial Intelligence › Speech Recognition and Synthesis
- Keywords
- scheduled sampling, credit assignment, greedy decoding, sequence-to-sequence models, exposure bias, named entity recognition, machine translation, sequence prediction
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.
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
