Papers by Tetsuro Morimura
Theoretical Guarantees for Minimum Bayes Risk Decoding (2025.acl-long)
Copied to clipboard
| Challenge: | Minimum Bayes Risk (MBR) decoding is a decision rule used to generate sequences from autoregressive probability models (e.g., LLMs). |
| Approach: | They propose to use minimum bayes risk (MBR) decoding to optimize output selection by maximizing expected utility value of an underlying human distribution. |
| Outcome: | The proposed method is effective even though the language space Y is larger than the hypothesis set. |
On the True Distribution Approximation of Minimum Bayes-Risk Decoding (2024.naacl-short)
Copied to clipboard
| Challenge: | Minimum Bayes-risk (MBR) decoding has recently gained renewed attention in text generation. |
| Approach: | They propose to use anomaly detection to measure the degree of approximation by sampling texts from a model and selecting the text with the highest similarity to the others. |
| Outcome: | The proposed method shows that previous hypotheses about samples do not correlate well with the variation, but the results support the core assumption of MBR decoding. |
Reinforcement Learning for Edit-Based Non-Autoregressive Neural Machine Translation (2024.naacl-srw)
Copied to clipboard
| Challenge: | Non-autoregressive (NAR) language models have a performance gap due to the large decoding space and difficulty in capturing dependency between target words accurately. |
| Approach: | They propose to use reinforcement learning to enhance the performance of edit-based NAR models by using stepwise reward maximization and episodic reward maximisation. |
| Outcome: | The proposed model outperforms autoregressive models in the evaluation of an edit-based model. |
Generating Diverse and High-Quality Texts by Minimum Bayes Risk Decoding (2024.findings-acl)
Copied to clipboard
| Challenge: | Existing decoding algorithms to generate diverse outputs are based on beam search or random sampling, thus their output quality is capped by these underlying decoding methods. |
| Approach: | They propose to add a diversity penalty to MBR decoding and a clustering problem to create diversity-promoting decoding algorithms by enforcing diversity objectives. |
| Outcome: | The proposed method achieves a better trade-off than the diverse beam search and sampling algorithms overall. |
Filtered Direct Preference Optimization (2024.emnlp-main)
Copied to clipboard
| Challenge: | Existing studies on the impact of RLHF on text quality have focused on reward-model-free RL. |
| Approach: | They propose an extension of direct preference optimization to improve model performance by analyzing the quality of the preference dataset. |
| Outcome: | The proposed method improves the performance of models optimized with DPO over those optimized with reward-model-based RLHF. |
Regularized Best-of-N Sampling with Minimum Bayes Risk Objective for Language Model Alignment (2025.naacl-long)
Copied to clipboard
| Challenge: | Best-of-N (BoN) sampling is an effective strategy for aligning Large Language Models (LLMs) to human preferences at the time of decoding. |
| Approach: | They propose a variant of BoN that incorporates the Minimum Bayes Risk objective as a proximity regularizer for BoN sampling. |
| Outcome: | The proposed method outperforms both BoN sampling and MBR decoding on the AlpacaFarm and Anthropic datasets. |