# Ghost in the Encoder: Decodable Artist Identity Representations in Lyrics-to-Song Generation

arXiv: https://arxiv.org/abs/2609.39552 · HTML: https://mlsanity.com/paper/2609.39552/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 30 of its 33 references that are in the map.
- First posted on arXiv 30 Sep 2026.

## Builds on (most cited first, 30 of 30)

- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Understanding intermediate layers using linear classifier probes](https://mlsanity.com/paper/1610.01644/) (2016)
- [Extracting Training Data from Diffusion Models](https://mlsanity.com/paper/2301.13188/) (2023)
- [Large-scale Contrastive Language-Audio Pretraining with Feature Fusion and Keyword-to-Caption Augmentation](https://mlsanity.com/paper/2211.06687/) (2022)
- [The Linear Representation Hypothesis and the Geometry of Large Language Models](https://mlsanity.com/paper/2311.03658/) (2023)
- [Jukebox: A Generative Model for Music](https://mlsanity.com/paper/2005.00341/) (2020)
- [Simple and Controllable Music Generation](https://mlsanity.com/paper/2306.05284/) (2023)
- [Scalable Extraction of Training Data from (Production) Language Models](https://mlsanity.com/paper/2311.17035/) (2023)
- [Diffusion Art or Digital Forgery? Investigating Data Replication in Diffusion Models](https://mlsanity.com/paper/2212.03860/) (2022)
- [Understanding and Mitigating Copying in Diffusion Models](https://mlsanity.com/paper/2305.20086/) (2023)
- [Long-form music generation with latent diffusion](https://mlsanity.com/paper/2404.10301/) (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)
- [DiffRhythm: Blazingly Fast and Embarrassingly Simple End-to-End Full-Length Song Generation with Latent Diffusion](https://mlsanity.com/paper/2503.01183/) (2025)
- [SongGen: A Single Stage Auto-regressive Transformer for Text-to-Song Generation](https://mlsanity.com/paper/2502.13128/) (2025)
- [ACE-Step 1.5: Pushing the Boundaries of Open-Source Music Generation](https://mlsanity.com/paper/2602.00744/) (2026)
- [SongBloom: Coherent Song Generation via Interleaved Autoregressive Sketching and Diffusion Refinement](https://mlsanity.com/paper/2506.07634/) (2025)
- [Exploring Musical Roots: Applying Audio Embeddings to Empower Influence Attribution for a Generative Music Model](https://mlsanity.com/paper/2401.14542/) (2024)
- [Do Music Generation Models Encode Music Theory?](https://mlsanity.com/paper/2410.00872/) (2024)
- [Discovering and Steering Interpretable Concepts in Large Generative Music Models](https://mlsanity.com/paper/2505.18186/) (2025)
- [SMITIN: Self-Monitored Inference-Time INtervention for Generative Music Transformers](https://mlsanity.com/paper/2404.02252/) (2024)
- [Large-Scale Training Data Attribution for Music Generative Models via Unlearning](https://mlsanity.com/paper/2506.18312/) (2025)
- [Supervised contrastive learning from weakly-labeled audio segments for musical version matching](https://mlsanity.com/paper/2502.16936/) (2025)
- [Temporal Concept Dynamics in Diffusion Models via Prompt-Conditioned Interventions](https://mlsanity.com/paper/2512.08486/) (2025)
- [TADA! Tuning Audio Diffusion Models through Activation Steering](https://mlsanity.com/paper/2602.11910/) (2026)
- [Bob's Confetti: Phonetic Memorization Attacks in Music and Video Generation](https://mlsanity.com/paper/2507.17937/) (2025)
- [Mitigating data replication in text-to-audio generative diffusion models through anti-memorization guidance](https://mlsanity.com/paper/2509.14934/) (2025)
- [The Name-Free Gap: Policy-Aware Stylistic Control in Music Generation](https://mlsanity.com/paper/2509.00654/) (2025)
- [The Artist is Present: Traces of Artists Resigind and Spawning in Text-to-Audio AI](https://mlsanity.com/paper/2511.17404/) (2025)

## Abstract (arXiv)

Text-to-song generation models can be prompted to imitate specific artists or regurgitate entire songs from their training data. Although these phenomena have been documented behaviorally on small datasets, little is known about the internal representations that may give rise to them. Prior interpretability work on generative audio has focused on locating semantic concepts such as genre or time signature within model activations. In this work, we show that a trained model can be probed for linearly decodable representations of artist identity from song lyrics alone, without any additional identifiers. Through a controlled case study of ACE-Step 1.5 spanning 2,000 songs across 100 artists, we demonstrate that the artist associated with a given set of lyrics can be identified within the model's internal activations, and that this conditioning signal propagates from the lyric encoder to the diffusion backbone during inference. These findings indicate that lyrics constitute an artist-level conditioning channel not addressed by prompt-side replication safeguards. More broadly, our work highlights how latent-space analysis can be used to audit what generative music models have implicitly learned from their training data.