# ECO Decoding: Entropy-Based Control for Controllability and Fluency in Controllable Dialogue Generation

arXiv: https://arxiv.org/abs/2511.01568 · HTML: https://mlsanity.com/paper/2511.01568/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 13 of its 23 references that are in the map.
- First posted on arXiv 3 Nov 2025.

## Builds on (most cited first, 13 of 13)

- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [A Diversity-Promoting Objective Function for Neural Conversation Models](https://mlsanity.com/paper/1510.03055/) (2015)
- [Neural Network Acceptability Judgments](https://mlsanity.com/paper/1805.12471/) (2018)
- [DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation](https://mlsanity.com/paper/1911.00536/) (2019)
- [DailyDialog: A Manually Labelled Multi-turn Dialogue Dataset](https://mlsanity.com/paper/1710.03957/) (2017)
- [DExperts: Decoding-Time Controlled Text Generation with Experts and Anti-Experts](https://mlsanity.com/paper/2105.03023/) (2021)
- [GeDi: Generative Discriminator Guided Sequence Generation](https://mlsanity.com/paper/2009.06367/) (2020)
- [FUDGE: Controlled Text Generation With Future Discriminators](https://mlsanity.com/paper/2104.05218/) (2021)
- [Deep Reinforcement Learning for Dialogue Generation](https://mlsanity.com/paper/1606.01541/) (2016)
- [MetaAligner: Towards Generalizable Multi-Objective Alignment of Language Models](https://mlsanity.com/paper/2403.17141/) (2024)
- [DIRECTOR: Generator-Classifiers For Supervised Language Modeling](https://mlsanity.com/paper/2206.07694/) (2022)
- [Seen to Unseen: Exploring Compositional Generalization of Multi-Attribute Controllable Dialogue Generation](https://mlsanity.com/paper/2306.10317/) (2023)

## Built on by (most cited first, 3 of 3)

- [IFHierBench: Hierarchical Instruction Following for Large Language Models](https://mlsanity.com/paper/2607.27912/) (2026)
- [Guidance Is Not a Hyperparameter: Learning Dynamic Control in Diffusion Language Models](https://mlsanity.com/paper/2605.07701/) (2026)
- [APCD: Adaptive Path-Contrastive Decoding for Reliable Large Language Model Generation](https://mlsanity.com/paper/2605.09492/) (2026)

## Abstract (arXiv)

Controllable Dialogue Generation (CDG) enables chatbots to generate responses with desired attributes, and weighted decoding methods have achieved significant success in the CDG task. However, using a fixed constant value to manage the bias of attribute probabilities makes it challenging to find an ideal control strength that satisfies both controllability and fluency. To address this issue, we propose ECO decoding (Entropy-based COntrol), which dynamically adjusts the control strength at each generation step according to the model's entropy in both the language model and attribute classifier probability distributions. Experiments on the DailyDialog and MultiWOZ datasets demonstrate that ECO decoding consistently improves controllability while maintaining fluency and grammaticality, outperforming prior decoding methods across various models and settings. Furthermore, ECO decoding alleviates probability interpolation issues in multi-attribute generation and consequently demonstrates strong performance in both single and multi-attribute scenarios.