# APEX: Large-scale Multi-task Aesthetic-Informed Popularity Prediction for AI-Generated Music

arXiv: https://arxiv.org/abs/2605.03395 · HTML: https://mlsanity.com/paper/2605.03395/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 21 of its 42 references that are in the map.
- First posted on arXiv 5 May 2026.

## Builds on (most cited first, 21 of 21)

- [Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics](https://mlsanity.com/paper/1705.07115/) (2017)
- [Simple and Controllable Music Generation](https://mlsanity.com/paper/2306.05284/) (2023)
- [Fr'echet Audio Distance: A Metric for Evaluating Music Enhancement Algorithms](https://mlsanity.com/paper/1812.08466/) (2018)
- [MERT: Acoustic Music Understanding Model with Large-Scale Self-supervised Training](https://mlsanity.com/paper/2306.00107/) (2023)
- [Meta Audiobox Aesthetics: Unified Automatic Quality Assessment for Speech, Music, and Sound](https://mlsanity.com/paper/2502.05139/) (2025)
- [ACE-Step: A Step Towards Music Generation Foundation Model](https://mlsanity.com/paper/2506.00045/) (2025)
- [SongEval: A Benchmark Dataset for Song Aesthetics Evaluation](https://mlsanity.com/paper/2505.10793/) (2025)
- [Music Arena: Live Evaluation for Text-to-Music](https://mlsanity.com/paper/2507.20900/) (2025)
- [A Comprehensive Survey for Evaluation Methodologies of AI-Generated Music](https://mlsanity.com/paper/2308.13736/) (2023)
- [A Survey on Evaluation Metrics for Music Generation](https://mlsanity.com/paper/2509.00051/) (2025)
- [MuQ-Eval: An Open-Source Per-Sample Quality Metric for AI Music Generation Evaluation](https://mlsanity.com/paper/2603.22677/) (2026)
- [Revisiting the problem of audio-based hit song prediction using convolutional neural networks](https://mlsanity.com/paper/1704.01280/) (2017)
- [Song Hit Prediction: Predicting Billboard Hits Using Spotify Data](https://mlsanity.com/paper/1908.08609/) (2019)
- [Dance Hit Song Prediction](https://mlsanity.com/paper/1905.08076/) (2019)
- [Universal Music Representations? Evaluating Foundation Models on World Music Corpora](https://mlsanity.com/paper/2506.17055/) (2025)
- [LSTM-RPA: A Simple but Effective Long Sequence Prediction Algorithm for Music Popularity Prediction](https://mlsanity.com/paper/2110.15790/) (2021)
- [An Analysis of Classification Approaches for Hit Song Prediction using Engineered Metadata Features with Lyrics and Audio Features](https://mlsanity.com/paper/2301.13507/) (2023)
- [Beyond Beats: A Recipe to Song Popularity? A machine learning approach](https://mlsanity.com/paper/2403.12079/) (2024)
- [Hit Song Prediction Based on Early Adopter Data and Audio Features](https://mlsanity.com/paper/2010.09489/) (2020)
- [Prediction of Spotify Chart Success Using Audio and Streaming Features](https://mlsanity.com/paper/2508.11632/) (2025)
- [Lyrics Matter: Exploiting the Power of Learnt Representations for Music Popularity Prediction](https://mlsanity.com/paper/2512.05508/) (2025)

## Abstract (arXiv)

Music popularity prediction has attracted growing research interest, with relevance to artists, platforms, and recommendation systems. However, the explosive rise of AI-generated music platforms has created an entirely new and largely unexplored landscape, where a surge of songs is produced and consumed daily without the traditional markers of artist reputation or label backing. Key, yet unexplored in this pursuit is aesthetic quality. We propose APEX, the first large-scale multi-task learning framework for AI-generated music, trained on over 211k songs (10k hours of audio) from Suno and Udio, that jointly predicts engagement-based popularity signals - streams and likes scores - alongside five perceptual aesthetic quality dimensions from frozen audio embeddings extracted from MERT, a self-supervised music understanding model. Aesthetic quality and popularity capture complementary aspects of music that together prove valuable: in an out-of-distribution evaluation on the Music Arena dataset, comprising pairwise human preference battles across eleven generative music systems unseen during training, including aesthetic features consistently improves preference prediction, demonstrating strong generalisation of the learned representations across generative architectures.