# Speech-Worthy Alignment for Japanese SpeechLLMs via Direct Preference Optimization

arXiv: https://arxiv.org/abs/2603.12565 · HTML: https://mlsanity.com/paper/2603.12565/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Preference alignment for recommenders and RAG, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 18 of its 21 references that are in the map.
- First posted on arXiv 13 Mar 2026.

## Placed in

- LLM alignment / RLHF / preference optimisation / [Preference alignment for recommenders and RAG](https://mlsanity.com/alignment/preference-alignment-for-recommenders-and-rag/)

## Builds on (most cited first, 18 of 18)

- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Robust Speech Recognition via Large-Scale Weak Supervision](https://mlsanity.com/paper/2212.04356/) (2022)
- [G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment](https://mlsanity.com/paper/2303.16634/) (2023)
- [Qwen2-Audio Technical Report](https://mlsanity.com/paper/2407.10759/) (2024)
- [SpeechGPT: Empowering Large Language Models with Intrinsic Cross-Modal Conversational Abilities](https://mlsanity.com/paper/2305.11000/) (2023)
- [SALMONN: Towards Generic Hearing Abilities for Large Language Models](https://mlsanity.com/paper/2310.13289/) (2023)
- [Listen, Think, and Understand](https://mlsanity.com/paper/2305.10790/) (2023)
- [VoiceBench: Benchmarking LLM-Based Voice Assistants](https://mlsanity.com/paper/2410.17196/) (2024)
- [LLaMA-Omni: Seamless Speech Interaction with Large Language Models](https://mlsanity.com/paper/2409.06666/) (2024)
- [GAMA: A Large Audio-Language Model with Advanced Audio Understanding and Complex Reasoning Abilities](https://mlsanity.com/paper/2406.11768/) (2024)
- [SLAM-Omni: Timbre-Controllable Voice Interaction System with Single-Stage Training](https://mlsanity.com/paper/2412.15649/) (2024)
- [Music Flamingo: Scaling Music Understanding in Audio Language Models](https://mlsanity.com/paper/2511.10289/) (2025)
- [OpenMU: Your Swiss Army Knife for Music Understanding](https://mlsanity.com/paper/2410.15573/) (2024)
- [DeepDialogue: A Multi-Turn Emotionally-Rich Spoken Dialogue Dataset](https://mlsanity.com/paper/2505.19978/) (2025)
- [Speechworthy Instruction-tuned Language Models](https://mlsanity.com/paper/2409.14672/) (2024)
- [DeepResonance: Enhancing Multimodal Music Understanding via Music-centric Multi-way Instruction Tuning](https://mlsanity.com/paper/2502.12623/) (2025)
- [Think, Verbalize, then Speak: Bridging Complex Thoughts and Comprehensible Speech](https://mlsanity.com/paper/2509.16028/) (2025)

## Abstract (arXiv)

SpeechLLMs typically combine ASR-trained encoders with text-based LLM backbones, leading them to inherit written-style output patterns unsuitable for text-to-speech synthesis. This mismatch is particularly pronounced in Japanese, where spoken and written registers differ substantially in politeness markers, sentence-final particles, and syntactic complexity. We propose a preference-based alignment approach to adapt Japanese SpeechLLMs for speech-worthy outputs: text that is concise, conversational, and readily synthesized as natural speech. To rigorously evaluate this task, we introduce SpokenElyza, a benchmark for Japanese speech-worthiness derived from ELYZA-tasks-100 with auditory verification by native experts. Experiments show that our approach achieves substantial improvement on SpokenElyza while largely preserving performance on the original written-style evaluation. We will release SpokenElyza to support future research on Japanese spoken dialog systems.