mlsanityData to 7 Oct 2026

It's MBR All the Way Down: Modern Generation Techniques Through the Lens of Minimum Bayes Risk

arXiv 2310.01387 ↗First posted 2 Oct 2023cs.CLcited by 49

Not on the map. Why?

Abstract (arXiv)

Minimum Bayes Risk (MBR) decoding is a method for choosing the outputs of a machine learning system based not on the output with the highest probability, but the output with the lowest risk (expected error) among multiple candidates. It is a simple but powerful method: for an additional cost at inference time, MBR provides reliable several-point improvements across metrics for a wide variety of tasks without any additional data or training. Despite this, MBR is not frequently applied in NLP works, and knowledge of the method itself is limited. We first provide an introduction to the method and the recent literature. We show that several recent methods that do not reference MBR can be written as special cases of MBR; this reformulation provides additional theoretical justification for the performance of these methods, explaining some results that were previously only empirical. We provide theoretical and empirical results about the effectiveness of various MBR variants and make concrete recommendations for the application of MBR in NLP models, including future directions in this area.

Key facts. Cited by 49 papers in the map (data through 7 Oct 2026). Builds on 44 of its 94 references that are in the map. First posted on arXiv 2 Oct 2023.

Builds on

44 of 94 references.

References · 44

  1. 2019BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehensioncited by 5,874
  2. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,601
  3. 2019BERTScore: Evaluating Text Generation with BERTcited by 5,462
  4. 2015Rethinking the Inception Architecture for Computer Visioncited by 5,451
  5. 2019The Curious Case of Neural Text Degenerationcited by 2,881
  6. 2018A Call for Clarity in Reporting BLEU Scorescited by 1,790
  7. 2022Competition-Level Code Generation with AlphaCodecited by 1,427
  8. 2014On the Properties of Neural Machine Translation: Encoder-Decoder Approachescited by 1,303
  9. 2020BLEURT: Learning Robust Metrics for Text Generationcited by 1,196
  10. 2018Averaging Weights Leads to Wider Optima and Better Generalizationcited by 1,194
  11. 2020Multilingual Denoising Pre-training for Neural Machine Translationcited by 1,028
  12. 2020COMET: A Neural Framework for MT Evaluationcited by 1,024
32 more
  1. 2016Abstractive Text Summarization Using Sequence-to-Sequence RNNs and Beyondcited by 809
  2. 2015Improving Neural Machine Translation Models with Monolingual Datacited by 784
  3. 2022Complexity-Based Prompting for Multi-Step Reasoningcited by 387
  4. 2017Six Challenges for Neural Machine Translationcited by 334
  5. 2020Is MAP Decoding All You Need? The Inadequacy of the Mode in Neural Machine Translationcited by 137
  6. 2022Natural Language to Code Translation with Executioncited by 122
  7. 2022Truncation Sampling as Language Model Desmoothingcited by 121
  8. 2018Correcting Length Bias in Neural Machine Translationcited by 119
  9. 2015Minimum Risk Training for Neural Machine Translationcited by 114
  10. 2019On NMT Search Errors and Model Errors: Cat Got Your Tongue?cited by 114
  11. 2021High Quality Rather than High Model Probability: Minimum Bayes Risk Decoding with Neural Metricscited by 104
  12. 2021Learning Compact Metrics for MTcited by 93
  13. 2022On Decoding Strategies for Neural Text Generatorscited by 89
  14. 2021Sampling-Based Approximations to Minimum Bayes Risk Decoding for Neural Machine Translationcited by 63
  15. 2023Epsilon Sampling Rocks: Investigating Sampling Strategies for Minimum Bayes Risk Decoding for Machine Translationcited by 57
  16. 2021Understanding the Properties of Minimum Bayes Risk Decoding in Neural Machine Translationcited by 56
  17. 2022Follow the Wisdom of the Crowd: Effective Text Generation via Minimum Bayes Risk Decodingcited by 54
  18. 2022Quality-Aware Decoding for Neural Machine Translationcited by 49
  19. 2022Identifying Weaknesses in Machine Translation Metrics Through Minimum Bayes Risk Decoding: A Case Study for COMETcited by 41
  20. 2021Maximum Bayes Smatch Ensemble Distillation for AMR Parsingcited by 27
  21. 2022Geographic Citation Gaps in NLP Researchcited by 19
  22. 2016Neural Machine Translation by Minimising the Bayes-risk with Respect to Syntactic Translation Latticescited by 17
  23. 2020Gender Gap in Natural Language Processing Research: Disparities in Authorship and Citationscited by 17
  24. 2019Leveraging Sentence Similarity in Natural Language Generation: Improving Beam Search using Range Votingcited by 15
  25. 2021Ensembling Graph Predictions for AMR Parsingcited by 15
  26. 2023Forgotten Knowledge: Examining the Citational Amnesia in NLPcited by 13
  27. 2016RIGA at SemEval-2016 Task 8: Impact of Smatch Extensions and Character-Level Neural Translation on AMR Parsing Accuracycited by 11
  28. 2023Lightweight reranking for language model generationscited by 10
  29. 2023CUED at ProbSum 2023: Hierarchical Ensemble of Summarization Modelscited by 8
  30. 2022DC-MBR: Distributional Cooling for Minimum Bayesian Risk Decodingcited by 4
  31. 2016Lexicons and Minimum Risk Training for Neural Machine Translation: NAIST-CMU at WAT2016cited by 3
  32. 2023AMRs Assemble! Learning to Ensemble with Autoregressive Models for AMR Parsingcited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 4; 2024: 22; 2025: 15; 2026: 8.

49 papers.

Sort:

By area:

  1. 2023A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questionscited by 1,582
  2. 2023Universal Self-Consistency for Large Language Model Generationcited by 198
  3. 2024From Decoding to Meta-Generation: Inference-time Algorithms for Large Language Modelscited by 136
  4. 2024DeAL: Decoding-time Alignment for Large Language Modelscited by 55
  5. 2025Evaluation of Best-of-N Sampling Strategies for Language Model Alignmentcited by 38
  6. 2023Faster Minimum Bayes Risk Decoding with Confidence-based Pruningcited by 34
  7. 2024Smoothie: Label Free Language Model Routingcited by 31
  8. 2024Regularized Best-of-N Sampling with Minimum Bayes Risk Objective for Language Model Alignmentcited by 24
  9. 2025Sample, Don't Search: Rethinking Test-Time Alignment for Language Modelscited by 21
  10. 2025Text Generation Beyond Discrete Token Samplingcited by 21
  11. 2024Better Instruction-Following Through Minimum Bayes Riskcited by 19
  12. 2024Narrowing the Knowledge Evaluation Gap: Open-Domain Question Answering with Multi-Granularity Answerscited by 18
37 more
  1. 2024Experts Don't Cheat: Learning What You Don't Know By Predicting Pairscited by 18
  2. 2024Hyperparameter-Free Approach for Faster Minimum Bayes Risk Decodingcited by 15
  3. 2024On Subjective Uncertainty Quantification and Calibration in Natural Language Generationcited by 14
  4. 2024Regression-aware Inference with LLMscited by 13
  5. 2024Improving Minimum Bayes Risk Decoding with Multi-Promptcited by 13
  6. 2024LIRE: listwise reward enhancement for preference alignmentcited by 11
  7. 2024REAL Sampling: Boosting Factuality and Diversity of Open-Ended Generation via Asymptotic Entropycited by 11
  8. 2023Model-Based Minimum Bayes Risk Decoding for Text Generationcited by 10
  9. 2024Diversity-Rewarded CFG Distillationcited by 9
  10. 2025When Life Gives You Samples: The Benefits of Scaling up Inference Compute for Multilingual LLMscited by 9
  11. 2024Merging Improves Self-Critique Against Jailbreak Attackscited by 8
  12. 2024Diversity Explains Inference Scaling Laws: Through a Case Study of Minimum Bayes Risk Decodingcited by 8
  13. 2024Compare without Despair: Reliable Preference Evaluation with Generation Separabilitycited by 4
  14. 2024Uncertainty-Guided Likelihood Tree Searchcited by 4
  15. 2025How Sampling Affects the Detectability of Machine-written texts: A Comprehensive Studycited by 4
  16. 2026Test-Time Scaling in Reasoning LLMs: Inference Regimes, Evaluation, and Reproducibilitycited by 4
  17. 2024Annotation-Efficient Language Model Alignment via Diverse and Representative Response Textscited by 3
  18. 2024Reranking Laws for Language Generation: A Communication-Theoretic Perspectivecited by 3
  19. 2024M-Ped: Multi-Prompt Ensemble Decoding for Large Language Modelscited by 3
  20. 2025Theoretical Guarantees for Minimum Bayes Risk Decodingcited by 3
  21. 2025Localized Calibrated Uncertainty in Code Language Modelscited by 3
  22. 2024On the True Distribution Approximation of Minimum Bayes-Risk Decodingcited by 2
  23. 2025Optimal Policy Minimum Bayesian Riskcited by 2
  24. 2026Task-Awareness Improves LLM Generations and Uncertaintycited by 2
  25. 2025CoKe: Customizable Fine-Grained Story Evaluation via Chain-of-Keyword Rationalizationcited by 1
  26. 2025Latent Beam Diffusion Models for Generating Visual Sequencescited by 1
  27. 2025Estimating Optimal Context Length for Hybrid Retrieval-augmented Multi-document Summarizationcited by 1
  28. 2025Learning Auxiliary Tasks Improves Reference-Free Hallucination Detection in Open-Domain Long-Form Generationcited by 1
  29. 2025Enhancing Temporal Awareness in LLMs for Temporal Point Processescited by 1
  30. 2025BRIDO: Bringing Democratic Order to Abstractive Summarizationcited by 0
  31. 2025Sample, Align, Synthesize: Graph-Based Response Synthesis with ConGrscited by 0
  32. 2026Consensus Group Relative Policy Optimization for Text Generationcited by 0
  33. 2026Clarify, Abstain or Answer? Strategising in Conversation with Belief-Augmented Generationcited by 0
  34. 2026Enhancing Factuality through Consensus and Consistency in Summarization Using Minimum Bayes Risk Decodingcited by 0
  35. 2026PPDL: LLM-Based Flows as Probabilistic Programscited by 0
  36. 2026Test-Time Scaling for CAD Generation via Verifier-Free Consensus Selectioncited by 0
  37. 2026Selective Regenerative Decoding: Trajectory-Level Intervention for Inference-Time Reasoningcited by 0