mlsanityData to 7 Oct 2026

Straight to the Gradient: Learning to Use Novel Tokens for Neural Text Generation

arXiv 2106.07207 ↗First posted 14 Jun 2021cs.CLcited by 19

Not on the map. Why?

Abstract (arXiv)

Advanced large-scale neural language models have led to significant success in many language generation tasks. However, the most commonly used training objective, Maximum Likelihood Estimation (MLE), has been shown problematic, where the trained model prefers using dull and repetitive phrases. In this work, we introduce ScaleGrad, a modification straight to the gradient of the loss function, to remedy the degeneration issue of the standard MLE objective. By directly maneuvering the gradient information, ScaleGrad makes the model learn to use novel tokens. Empirical results show the effectiveness of our method not only in open-ended generation, but also in directed generation tasks. With the simplicity in architecture, our method can serve as a general training objective that is applicable to most of the neural text generation tasks.

Key facts. Cited by 19 papers in the map (data through 7 Oct 2026). Builds on 29 of its 43 references that are in the map. First posted on arXiv 14 Jun 2021.

Builds on

29 of 43 references.

References · 29

  1. 2017Attention Is All You Needcited by 59,932
  2. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  3. 2020Language Models are Few-Shot Learnerscited by 32,199
  4. 2019BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehensioncited by 5,874
  5. 2016Pointer Sentinel Mixture Modelscited by 2,992
  6. 2015Neural Machine Translation of Rare Words with Subword Unitscited by 2,898
  7. 2019The Curious Case of Neural Text Degenerationcited by 2,881
  8. 2014CIDEr: Consensus-based Image Description Evaluationcited by 2,027
  9. 2016Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translationcited by 1,471
  10. 2017Get To The Point: Summarization with Pointer-Generator Networkscited by 1,281
  11. 2018Hierarchical Neural Story Generationcited by 1,182
  12. 2015Teaching Machines to Read and Comprehendcited by 1,039
17 more
  1. 2016Abstractive Text Summarization Using Sequence-to-Sequence RNNs and Beyondcited by 809
  2. 2019Text Summarization with Pretrained Encoderscited by 553
  3. 2019Neural Text Generation with Unlikelihood Trainingcited by 489
  4. 2019MoverScore: Text Generation Evaluating with Contextualized Embeddings and Earth Mover Distancecited by 415
  5. 2017A Deep Reinforced Model for Abstractive Summarizationcited by 364
  6. 2017Six Challenges for Neural Machine Translationcited by 334
  7. 2020ProphetNet: Predicting Future N-gram for Sequence-to-Sequence Pre-trainingcited by 221
  8. 2019Stochastic Beams and Where to Find Them: The Gumbel-Top-k Trick for Sampling Sequences Without Replacementcited by 205
  9. 2019On NMT Search Errors and Model Errors: Cat Got Your Tongue?cited by 114
  10. 2016Modeling Coverage for Neural Machine Translationcited by 110
  11. 2019Don't Say That! Making Inconsistent Dialogue Unlikely with Unlikelihood Trainingcited by 110
  12. 2016A Hierarchical Approach for Generating Descriptive Image Paragraphscited by 104
  13. 2019Describing like humans: on diversity in image captioningcited by 46
  14. 2016Learning-Based Single-Document Summarization with Compression and Anaphoricity Constraintscited by 34
  15. 2020Contextual Temperature for Language Modelingcited by 25
  16. 2020TLDR: Token Loss Dynamic Reweighting for Reducing Repetitive Utterance Generationcited by 6
  17. 2021Improving Diversity of Neural Text Generation via Inverse Probability Weightingcited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 4; 2023: 9; 2024: 2; 2025: 2; 2026: 2.

19 papers.

Sort:

By area:

  1. 2022Learning to Break the Loop: Analyzing and Mitigating Repetitions for Neural Text Generationcited by 90
  2. 2025Jointly Reinforcing Diversity and Quality in Language Model Generationscited by 79
  3. 2022SummaReranker: A Multi-Task Mixture-of-Experts Re-ranking Framework for Abstractive Summarizationcited by 61
  4. 2023Repetition In Repetition Out: Towards Understanding Neural Text Degeneration from the Data Perspectivecited by 39
  5. 2024From Self-Attention to Markov Models: Unveiling the Dynamics of Generative Transformerscited by 34
  6. 2023Understanding In-Context Learning from Repetitionscited by 27
  7. 2025Unveiling Attractor Cycles in Large Language Models: A Dynamical Systems View of Successive Paraphrasingcited by 14
  8. 2024Adversarial Representation Engineering: A General Model Editing Framework for Large Language Modelscited by 12
  9. 2022Nearest Neighbor Knowledge Distillation for Neural Machine Translationcited by 10
  10. 2022R2D2: Robust Data-to-Text with Replacement Detectioncited by 8
  11. 2023Language Model Decoding as Direct Metrics Optimizationcited by 8
  12. 2023Error Norm Truncation: Robust Training in the Presence of Data Noise for Text Generation Modelscited by 6
7 more
  1. 2023Multi-level Adaptive Contrastive Learning for Knowledge Internalization in Dialogue Generationcited by 4
  2. 2023Decouple Non-parametric Knowledge Distillation For End-to-end Speech Translationcited by 3
  3. 2023Mitigating the Learning Bias towards Repetition by Self-Contrastive Training for Open-Ended Generationcited by 3
  4. 2026Resource Consumption Threats in Large Language Modelscited by 2
  5. 2023Dynamic Scheduled Sampling with Imitation Loss for Neural Text Generationcited by 0
  6. 2023Exploring Automatic Text Simplification of German Narrative Documentscited by 0
  7. 2026FOCUS & RePAIR: Mitigating Text Degeneration via Token-Level Guidance for Pruned Large Language Modelscited by 0