# ConceptCaps: a Distilled Concept Dataset for Interpretability in Music Models

arXiv: https://arxiv.org/abs/2601.14157 · HTML: https://mlsanity.com/paper/2601.14157/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 18 of its 31 references that are in the map.
- First posted on arXiv 20 Jan 2026.

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

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [BERTScore: Evaluating Text Generation with BERT](https://mlsanity.com/paper/1904.09675/) (2019)
- [Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)](https://mlsanity.com/paper/1711.11279/) (2017)
- [StackGAN: Text to Photo-realistic Image Synthesis with Stacked Generative Adversarial Networks](https://mlsanity.com/paper/1612.03242/) (2016)
- [Simple and Controllable Music Generation](https://mlsanity.com/paper/2306.05284/) (2023)
- [MAUVE: Measuring the Gap Between Neural Text and Human Text using Divergence Frontiers](https://mlsanity.com/paper/2102.01454/) (2021)
- [MusicLM: Generating Music From Text](https://mlsanity.com/paper/2301.11325/) (2023)
- [WavCaps: A ChatGPT-Assisted Weakly-Labelled Audio Captioning Dataset for Audio-Language Multimodal Research](https://mlsanity.com/paper/2303.17395/) (2023)
- [Adapting Frechet Audio Distance for Generative Music Evaluation](https://mlsanity.com/paper/2311.01616/) (2023)
- [LP-MusicCaps: LLM-Based Pseudo Music Captioning](https://mlsanity.com/paper/2307.16372/) (2023)
- [The Song Describer Dataset: a Corpus of Audio Captions for Music-and-Language Evaluation](https://mlsanity.com/paper/2311.10057/) (2023)
- [Sound-VECaps: Improving Audio Generation with Visual Enhanced Captions](https://mlsanity.com/paper/2407.04416/) (2024)
- [Concept-Based Techniques for "Musicologist-friendly" Explanations in a Deep Music Classifier](https://mlsanity.com/paper/2208.12485/) (2022)
- [MusGO: A Community-Driven Framework For Assessing Openness in Music-Generative AI](https://mlsanity.com/paper/2507.03599/) (2025)
- [Human-CLAP: Human-perception-based contrastive language-audio pretraining](https://mlsanity.com/paper/2506.23553/) (2025)
- [Learning Unsupervised Hierarchies of Audio Concepts](https://mlsanity.com/paper/2207.11231/) (2022)
- [Beyond Genre: Diagnosing Bias in Music Embeddings Using Concept Activation Vectors](https://mlsanity.com/paper/2509.24482/) (2025)

## Abstract (arXiv)

Concept-based interpretability methods like TCAV require clean, well-separated positive and negative examples for each concept. Existing music datasets lack this structure: tags are sparse, noisy, or ill-defined. We introduce ConceptCaps, a dataset of 21k music-caption-tags triplets with explicit labels from a 200-attribute taxonomy. Our pipeline separates semantic modeling from text generation: a VAE learns plausible attribute co-occurrence patterns, a fine-tuned LLM converts attribute lists into professional descriptions, and MusicGen synthesizes corresponding audio. This separation improves coherence and controllability over end-to-end approaches. We validate the dataset through audio-text alignment (CLAP), linguistic quality metrics (BERTScore, MAUVE), and TCAV analysis confirming that concept probes recover musically meaningful patterns. Dataset and code are available online.