# CMI-RewardBench: Evaluating Music Reward Models with Compositional Multimodal Instruction

arXiv: https://arxiv.org/abs/2603.00610 · HTML: https://mlsanity.com/paper/2603.00610/

## Key facts

- Cited by 5 papers in the map (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 53 of its 70 references that are in the map.
- First posted on arXiv 28 Feb 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, 30 of 53)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities](https://mlsanity.com/paper/2507.06261/) (2025)
- [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2204.05862/) (2022)
- [Learning to summarize from human feedback](https://mlsanity.com/paper/2009.01325/) (2020)
- [A Survey on LLM-as-a-Judge](https://mlsanity.com/paper/2411.15594/) (2024)
- [Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preference](https://mlsanity.com/paper/2403.04132/) (2024)
- [ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generation](https://mlsanity.com/paper/2304.05977/) (2023)
- [Pick-a-Pic: An Open Dataset of User Preferences for Text-to-Image Generation](https://mlsanity.com/paper/2305.01569/) (2023)
- [Qwen2.5-Omni Technical Report](https://mlsanity.com/paper/2503.20215/) (2025)
- [Large-scale Contrastive Language-Audio Pretraining with Feature Fusion and Keyword-to-Caption Augmentation](https://mlsanity.com/paper/2211.06687/) (2022)
- [AudioLDM: Text-to-Audio Generation with Latent Diffusion Models](https://mlsanity.com/paper/2301.12503/) (2023)
- [Qwen2-Audio Technical Report](https://mlsanity.com/paper/2407.10759/) (2024)
- [Simple and Controllable Music Generation](https://mlsanity.com/paper/2306.05284/) (2023)
- [MusicLM: Generating Music From Text](https://mlsanity.com/paper/2301.11325/) (2023)
- [AudioLDM 2: Learning Holistic Audio Generation with Self-supervised Pretraining](https://mlsanity.com/paper/2308.05734/) (2023)
- [Improving Video Generation with Human Feedback](https://mlsanity.com/paper/2501.13918/) (2025)
- [Stable Audio Open](https://mlsanity.com/paper/2407.14358/) (2024)
- [HIVE: Harnessing Human Feedback for Instructional Visual Editing](https://mlsanity.com/paper/2303.09618/) (2023)
- [MusicLDM: Enhancing Novelty in Text-to-Music Generation Using Beat-Synchronous Mixup Strategies](https://mlsanity.com/paper/2308.01546/) (2023)
- [GenAI Arena: An Open Evaluation Platform for Generative Models](https://mlsanity.com/paper/2406.04485/) (2024)
- [YuE: Scaling Open Foundation Models for Long-Form Music Generation](https://mlsanity.com/paper/2503.08638/) (2025)
- [MuQ: Self-Supervised Music Representation Learning with Mel Residual Vector Quantization](https://mlsanity.com/paper/2501.01108/) (2025)
- [ACE-Step: A Step Towards Music Generation Foundation Model](https://mlsanity.com/paper/2506.00045/) (2025)
- [Read, Watch and Scream! Sound Generation from Text and Video](https://mlsanity.com/paper/2407.05551/) (2024)
- [SpeechAlign: Aligning Speech Generation to Human Preferences](https://mlsanity.com/paper/2404.05600/) (2024)
- [DiffRhythm: Blazingly Fast and Embarrassingly Simple End-to-End Full-Length Song Generation with Latent Diffusion](https://mlsanity.com/paper/2503.01183/) (2025)
- [Foundation Models for Music: A Survey](https://mlsanity.com/paper/2408.14340/) (2024)
- [Tuning Large Multimodal Models for Videos using Reinforcement Learning from AI Feedback](https://mlsanity.com/paper/2402.03746/) (2024)
- [MusicRL: Aligning Music Generation to Human Preferences](https://mlsanity.com/paper/2402.04229/) (2024)

## Built on by (most cited first, 5 of 5)

- [TuneJury: An Open Metric for Improving Music Generation Preference Alignment](https://mlsanity.com/paper/2606.17006/) (2026)
- [Pushing the Frontier of Full-Song Generation: Hierarchical Autoregressive Planning Meets Flow-Matching Rendering](https://mlsanity.com/paper/2607.20253/) (2026)
- [MuseCritic: Learning Multi-Aspect Song Rewards through Natural-Language Aesthetic Critiques](https://mlsanity.com/paper/2608.11755/) (2026)
- [VIBE: Video Instruction-aligned Background music gEneration](https://mlsanity.com/paper/2608.30125/) (2026)
- [MIRA: A Musical Intent Refinement Agent for Aligning Text-to-Music Generation with User Intent](https://mlsanity.com/paper/2610.10355/) (2026)

## Abstract (arXiv)

While music generation models have evolved to handle complex multimodal inputs mixing text, lyrics, and reference audio, evaluation mechanisms have lagged behind. In this paper, we bridge this critical gap by establishing a comprehensive ecosystem for music reward modeling under Compositional Multimodal Instruction (CMI), where the generated music may be conditioned on text descriptions, lyrics, and audio prompts. We first introduce CMI-Pref-Pseudo, a large-scale preference dataset comprising 110k pseudo-labeled samples, and CMI-Pref, a high-quality, human-annotated corpus tailored for fine-grained alignment tasks. To unify the evaluation landscape, we propose CMI-RewardBench, a unified benchmark that evaluates music reward models on heterogeneous samples across musicality, text-music alignment, and compositional instruction alignment. Leveraging these resources, we develop CMI reward models (CMI-RMs), a parameter-efficient reward model family capable of processing heterogeneous inputs. We evaluate their correlation with human judgment scores on musicality and alignment on CMI-Pref along with previous datasets. Further experiments demonstrate that CMI-RM not only correlates strongly with human judgments, but also enables effective inference-time scaling via top-k filtering. Code is available at GitHub (https://github.com/Haiwen-Xia/CMI-RewardBench). Model weights: CMI-RM (https://huggingface.co/HaiwenXia/CMI-RM). Datasets: CMI-Pref-Pseudo (https://huggingface.co/datasets/HaiwenXia/cmi-pref-pseudo) and CMI-Pref (https://huggingface.co/datasets/HaiwenXia/cmi-pref)