CGMH: Constrained Sentence Generation by Metropolis-Hastings Sampling
Not on the map. Why?
Abstract (arXiv)
In real-world applications of natural language generation, there are often constraints on the target sentences in addition to fluency and naturalness requirements. Existing language generation techniques are usually based on recurrent neural networks (RNNs). However, it is non-trivial to impose constraints on RNNs while maintaining generation quality, since RNNs generate sentences sequentially (or with beam search) from the first word to the last. In this paper, we propose CGMH, a novel approach using Metropolis-Hastings sampling for constrained sentence generation. CGMH allows complicated constraints such as the occurrence of multiple keywords in the target sentences, which cannot be handled in traditional RNN-based approaches. Moreover, CGMH works in the inference stage, and does not require parallel corpora for training. We evaluate our method on a variety of tasks, including keywords-to-sentence generation, unsupervised sentence paraphrasing, and unsupervised sentence error correction. CGMH achieves high performance compared with previous supervised methods for sentence generation. Our code is released at https://github.com/NingMiao/CGMH
Key facts. Cited by 94 papers in the map (data through 14 Oct 2026). First posted on arXiv 14 Nov 2018.
Builds on
No reference list yet.
Built on by
Citing papers per year: 2021: 28; 2022: 19; 2023: 14; 2024: 10; 2025: 13; 2026: 10.
94 papers.
Sort:
By area:
- 2023Analyzing Leakage of Personally Identifiable Information in Language Modelscited by 285
- 2022Generating Sequences by Learning to Self-Correctcited by 277
- 2022COLD Decoding: Energy-based Constrained Text Generation with Langevin Dynamicscited by 161
- 2021NeuroLogic A*esque Decoding: Constrained Text Generation with Lookahead Heuristicscited by 154
- 2021Recent Advances in Deep Learning Based Dialogue Systems: A Systematic Surveycited by 130
- 2021MARS: Markov Molecular Sampling for Multi-objective Drug Discoverycited by 125
- 2023Aligning Language Models with Preferences through f-divergence Minimizationcited by 120
- 2023Amortizing intractable inference in large language modelscited by 97
- 2021Improving Coherence and Consistency in Neural Sequence Models with Dual-System, Neuro-Symbolic Reasoningcited by 85
- 2023Sequential Monte Carlo Steering of Large Language Models using Probabilistic Programscited by 80
- 2021A Plug-and-Play Method for Controlled Text Generationcited by 56
- 2025Edit Flows: Flow Matching with Edit Operationscited by 55
38 more
- 2025Syntactic and Semantic Control of Large Language Models via Sequential Monte Carlocited by 46
- 2024Space Group Informed Transformer for Crystalline Materials Generationcited by 41
- 2021Coherence boosting: When your pretrained language model is not paying enough attentioncited by 38
- 2024Automatic Pseudo-Harmful Prompt Generation for Evaluating False Refusals in Large Language Modelscited by 35
- 2023Impossible Distillation: from Low-Quality Model to High-Quality Dataset & Model for Summarization and Paraphrasingcited by 31
- 2022Learning to Model Editing Processescited by 30
- 2022Faithfulness in Natural Language Generation: A Systematic Survey of Analysis, Evaluation and Optimization Methodscited by 27
- 2021Parallel Refinements for Lexically Constrained Text Generation with BARTcited by 24
- 2025Sample, Don't Search: Rethinking Test-Time Alignment for Language Modelscited by 22
- 2022Why is constrained neural language generation particularly challenging?cited by 19
- 2021Controllable Summarization with Constrained Markov Decision Processcited by 17
- 2022Text Revision by On-the-Fly Representation Optimizationcited by 17
- 2025Constrained Discrete Diffusioncited by 17
- 2023Prompt-Based Editing for Text Style Transfercited by 16
- 2024QUEST: Quality-Aware Metropolis-Hastings Sampling for Machine Translationcited by 16
- 2021NAREOR: The Narrative Reordering Problemcited by 14
- 2023Zero-shot Clarifying Question Generation for Conversational Searchcited by 14
- 2021ConRPG: Paraphrase Generation using Contexts as Regularizercited by 13
- 2022A Character-Level Length-Control Algorithm for Non-Autoregressive Sentence Summarizationcited by 13
- 2023Don't Fine-Tune, Decode: Syntax Error-Free Tool Use via Constrained Decodingcited by 13
- 2023PathFinder: Guided Search over Multi-Step Reasoning Pathscited by 12
- 2025Flexible-length Text Infilling for Discrete Diffusion Modelscited by 12
- 2022Teacher Forcing Recovers Reward Functions for Text Generationcited by 11
- 2022Quantum Natural Language Generation on Near-Term Devicescited by 11
- 2021Unsupervised Contextual Paraphrase Generation using Lexical Control and Reinforcement Learningcited by 10
- 2021Show Me How To Revise: Improving Lexically Constrained Sentence Generation with XLNetcited by 10
- 2021Retrieve, Caption, Generate: Visual Grounding for Enhancing Commonsense in Text Generation Modelscited by 9
- 2022Converge to the Truth: Factual Error Correction via Iterative Constrained Editingcited by 9
- 2024Latent Logic Tree Extraction for Event Sequence Explanation from LLMscited by 9
- 2021SAPPHIRE: Approaches for Enhanced Concept-to-Text Generationcited by 8
- 2021Unsupervised Editing for Counterfactual Storiescited by 8
- 2024Constrained Decoding for Cross-lingual Label Projectioncited by 8
- 2021Language Model Augmented Relevance Scorecited by 7
- 2025Neuro-Symbolic Predictive Process Monitoringcited by 7
- 2021Controlling Text Edition by Changing Answers of Specific Questionscited by 6
- 2022UCEpic: Unifying Aspect Planning and Lexical Constraints for Generating Explanations in Recommendationcited by 6
- 2022AutoTemplate: A Simple Recipe for Lexically Constrained Text Generationcited by 6
- 2021Triangular Bidword Generation for Sponsored Search Auctioncited by 5
50 most cited of 94 All 94 (JSON)
- 2026REWIND: Uncertainty-Guided Trajectory Repair for Test-Time Reasoning14 Oct 2026
- 2026Making Grid Beam Search Less Greedy30 Sep 2026
- 2026BayesPrompt: human readable prompts that make sense18 Aug 2026
- 2026Verbalized Particle Posterior: Bayesian Inference over Natural Language Hypotheses25 Jul 2026
- 2026Structured Inference with Large Language Gibbs17 Jun 2026
- 2026Proximal Policy Optimization for Amortized Discrete Sampling14 Jun 2026
- 2026Neuro-Symbolic Injection of LTLf Constraints in Autoregressive Reinforcement Learning Policies6 Jun 2026
- 2026CellxPert: Inference-Time MCMC Steering of a Multi-Omics Single-Cell Foundation Model for In-Silico Perturbation30 Apr 2026
- 2026Length Value Model: Scalable Value Pretraining for Token-Level Length Modeling29 Apr 2026
- 2026Ensembling Language Models with Sequential Monte Carlo5 Mar 2026
- 2025STARS: Synchronous Token Alignment for Robust Supervision in Large Language Models5 Nov 2025
- 2025Neuro-Symbolic Predictive Process Monitoring31 Aug 2025
38 more
- 2025SMCLM: Semantically Meaningful Causal Language Modeling for Autoregressive Paraphrase Generation4 Jul 2025
- 2025Flexible-length Text Infilling for Discrete Diffusion Models16 Jun 2025
- 2025ABS: Enforcing Constraint Satisfaction On Generated Sequences Via Automata-Guided Beam Search11 Jun 2025
- 2025Edit Flows: Flow Matching with Edit Operations10 Jun 2025
- 2025Semantic Probabilistic Control of Language Models4 May 2025
- 2025Syntactic and Semantic Control of Large Language Models via Sequential Monte Carlo17 Apr 2025
- 2025Efficient and Asymptotically Unbiased Constrained Decoding for Large Language Models12 Apr 2025
- 2025Sample, Don't Search: Rethinking Test-Time Alignment for Language Models4 Apr 2025
- 2025QIRL: Optimized Question-Image Relation Learning for Bias-Robust Visual Question Answering4 Apr 2025
- 2025Constrained Discrete Diffusion12 Mar 2025
- 2025Fake News Detection After LLM Laundering: Measurement and Explanation29 Jan 2025
- 2024Learning to Adapt to Low-Resource Paraphrase Generation22 Dec 2024
- 2024Local and Global Decoding in Text Generation14 Oct 2024
- 2024Edit-Constrained Decoding for Sentence Simplification28 Sep 2024
- 2024Automatic Pseudo-Harmful Prompt Generation for Evaluating False Refusals in Large Language Models1 Sep 2024
- 2024Latent Logic Tree Extraction for Event Sequence Explanation from LLMs3 Jun 2024
- 2024QUEST: Quality-Aware Metropolis-Hastings Sampling for Machine Translation28 May 2024
- 2024Effective Unsupervised Constrained Text Generation based on Perturbed Masking24 Apr 2024
- 2024Space Group Informed Transformer for Crystalline Materials Generation23 Mar 2024
- 2024Multi-Intent Attribute-Aware Text Matching in Searching12 Feb 2024
- 2024Constrained Decoding for Cross-lingual Label Projection5 Feb 2024
- 2023PathFinder: Guided Search over Multi-Step Reasoning Paths8 Dec 2023
- 2023Evaluating, Understanding, and Improving Constrained Text Generation for Large Language Models25 Oct 2023
- 2023DiQAD: A Benchmark Dataset for End-to-End Open-domain Dialogue Assessment25 Oct 2023
- 2023Paraphrase Types for Generation and Detection23 Oct 2023
- 2023Don't Fine-Tune, Decode: Syntax Error-Free Tool Use via Constrained Decoding10 Oct 2023
- 2023Amortizing intractable inference in large language models6 Oct 2023
- 2023Search and Learning for Unsupervised Text Generation18 Sep 2023
- 2023On the Depth between Beam Search and Exhaustive Search for Text Generation25 Aug 2023
- 2023Sequential Monte Carlo Steering of Large Language Models using Probabilistic Programs5 Jun 2023
- 2023Impossible Distillation: from Low-Quality Model to High-Quality Dataset & Model for Summarization and Paraphrasing26 May 2023
- 2023Aligning Language Models with Preferences through f-divergence Minimization16 Feb 2023
- 2023Analyzing Leakage of Personally Identifiable Information in Language Models1 Feb 2023
- 2023Zero-shot Clarifying Question Generation for Conversational Search30 Jan 2023
- 2023Prompt-Based Editing for Text Style Transfer27 Jan 2023
- 2022Harnessing Knowledge and Reasoning for Human-Like Natural Language Generation: A Brief Review7 Dec 2022
- 2022Lexical Complexity Controlled Sentence Generation26 Nov 2022
- 2022Converge to the Truth: Factual Error Correction via Iterative Constrained Editing22 Nov 2022
50 newest of 94 All 94 (JSON)
- 2025Edit Flows: Flow Matching with Edit Operationscited by 55
- 2025Constrained Discrete Diffusioncited by 17
- 2025Flexible-length Text Infilling for Discrete Diffusion Modelscited by 12
- 2025Flexible-length Text Infilling for Discrete Diffusion Models16 Jun 2025
- 2025Edit Flows: Flow Matching with Edit Operations10 Jun 2025
- 2025Constrained Discrete Diffusion12 Mar 2025
- 2026Proximal Policy Optimization for Amortized Discrete Sampling14 Jun 2026
- 2022Generating Sequences by Learning to Self-Correctcited by 277
- 2022COLD Decoding: Energy-based Constrained Text Generation with Langevin Dynamicscited by 161
- 2021NeuroLogic A*esque Decoding: Constrained Text Generation with Lookahead Heuristicscited by 154
- 2021Recent Advances in Deep Learning Based Dialogue Systems: A Systematic Surveycited by 130
- 2021MARS: Markov Molecular Sampling for Multi-objective Drug Discoverycited by 125
- 2023Amortizing intractable inference in large language modelscited by 97
- 2021Improving Coherence and Consistency in Neural Sequence Models with Dual-System, Neuro-Symbolic Reasoningcited by 85
- 2023Sequential Monte Carlo Steering of Large Language Models using Probabilistic Programscited by 80
- 2021A Plug-and-Play Method for Controlled Text Generationcited by 56
- 2025Syntactic and Semantic Control of Large Language Models via Sequential Monte Carlocited by 46
- 2024Space Group Informed Transformer for Crystalline Materials Generationcited by 41
- 2021Coherence boosting: When your pretrained language model is not paying enough attentioncited by 38
12 most cited of 81 All 81 (JSON)
- 2026Making Grid Beam Search Less Greedy30 Sep 2026
- 2026BayesPrompt: human readable prompts that make sense18 Aug 2026
- 2026Verbalized Particle Posterior: Bayesian Inference over Natural Language Hypotheses25 Jul 2026
- 2026Structured Inference with Large Language Gibbs17 Jun 2026
- 2026CellxPert: Inference-Time MCMC Steering of a Multi-Omics Single-Cell Foundation Model for In-Silico Perturbation30 Apr 2026
- 2026Length Value Model: Scalable Value Pretraining for Token-Level Length Modeling29 Apr 2026
- 2026Ensembling Language Models with Sequential Monte Carlo5 Mar 2026
- 2025Neuro-Symbolic Predictive Process Monitoring31 Aug 2025
- 2025SMCLM: Semantically Meaningful Causal Language Modeling for Autoregressive Paraphrase Generation4 Jul 2025
- 2025ABS: Enforcing Constraint Satisfaction On Generated Sequences Via Automata-Guided Beam Search11 Jun 2025
- 2025Semantic Probabilistic Control of Language Models4 May 2025
- 2025Syntactic and Semantic Control of Large Language Models via Sequential Monte Carlo17 Apr 2025
12 newest of 81 All 81 (JSON)