# Negative Training for Neural Dialogue Response Generation

arXiv: https://arxiv.org/abs/1903.02134 · HTML: https://mlsanity.com/paper/1903.02134/

## Key facts

- Cited by 24 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 6 Mar 2019.

## Built on by (most cited first, 24 of 24)

- [Red Teaming Language Models with Language Models](https://mlsanity.com/paper/2202.03286/) (2022)
- [Recent Advances in Deep Learning Based Dialogue Systems: A Systematic Survey](https://mlsanity.com/paper/2105.04387/) (2021)
- [CLIFF: Contrastive Learning for Improving Faithfulness and Factuality in Abstractive Summarization](https://mlsanity.com/paper/2109.09209/) (2021)
- [The CRINGE Loss: Learning what language not to model](https://mlsanity.com/paper/2211.05826/) (2022)
- [Rethinking and Refining the Distinct Metric](https://mlsanity.com/paper/2202.13587/) (2022)
- [ConTinTin: Continual Learning from Task Instructions](https://mlsanity.com/paper/2203.08512/) (2022)
- [Diversifying Dialog Generation via Adaptive Label Smoothing](https://mlsanity.com/paper/2105.14556/) (2021)
- [Evolving Diverse Red-team Language Models in Multi-round Multi-agent Games](https://mlsanity.com/paper/2310.00322/) (2023)
- [Measuring and Improving Semantic Diversity of Dialogue Generation](https://mlsanity.com/paper/2210.05725/) (2022)
- [Identifying Untrustworthy Samples: Data Filtering for Open-domain Dialogues with Bayesian Optimization](https://mlsanity.com/paper/2109.06471/) (2021)
- [Diversifying Neural Dialogue Generation via Negative Distillation](https://mlsanity.com/paper/2205.02795/) (2022)
- [Automatically Exposing Problems with Neural Dialog Models](https://mlsanity.com/paper/2109.06950/) (2021)
- [Unified Multimodal Model with Unlikelihood Training for Visual Dialog](https://mlsanity.com/paper/2211.13235/) (2022)
- [Controlling the Focus of Pretrained Language Generation Models](https://mlsanity.com/paper/2203.01146/) (2022)
- [MOCHA: A Multi-Task Training Approach for Coherent Text Generation from Cognitive Perspective](https://mlsanity.com/paper/2210.14650/) (2022)
- [Learning Descriptive Image Captioning via Semipermeable Maximum Likelihood Estimation](https://mlsanity.com/paper/2306.13460/) (2023)
- [IdealTSF: Can Non-Ideal Data Contribute to Enhancing the Performance of Time Series Forecasting Models?](https://mlsanity.com/paper/2512.05442/) (2025)
- [Diversifying Neural Text Generation with Part-of-Speech Guided Softmax and Sampling](https://mlsanity.com/paper/2105.03641/) (2021)
- [Similarity Based Label Smoothing For Dialogue Generation](https://mlsanity.com/paper/2107.11481/) (2021)
- [Near-Negative Distinction: Giving a Second Life to Human Evaluation Datasets](https://mlsanity.com/paper/2205.06871/) (2022)
- [Modeling Complex Dialogue Mappings via Sentence Semantic Segmentation Guided Conditional Variational Auto-Encoder](https://mlsanity.com/paper/2212.00231/) (2022)
- [Towards Minimal Targeted Updates of Language Models with Targeted Negative Training](https://mlsanity.com/paper/2406.13660/) (2024)
- [Learning to Diversify Neural Text Generation via Degenerative Model](https://mlsanity.com/paper/2309.12619/) (2023)
- [Reducing the Probability of Undesirable Outputs in Language Models Using Probabilistic Inference](https://mlsanity.com/paper/2510.21184/) (2025)

## Abstract (arXiv)

Although deep learning models have brought tremendous advancements to the field of open-domain dialogue response generation, recent research results have revealed that the trained models have undesirable generation behaviors, such as malicious responses and generic (boring) responses. In this work, we propose a framework named "Negative Training" to minimize such behaviors. Given a trained model, the framework will first find generated samples that exhibit the undesirable behavior, and then use them to feed negative training signals for fine-tuning the model. Our experiments show that negative training can significantly reduce the hit rate of malicious responses, or discourage frequent responses and improve response diversity.