REVIEW 32 cited by
Music Transformer
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Music relies heavily on repetition to build structure and meaning. Self-reference occurs on multiple timescales, from motifs to phrases to reusing of entire sections of music, such as in pieces with ABA structure. The Transformer (Vaswani et al., 2017), a sequence model based on self-attention, has achieved compelling results in many generation tasks that require maintaining long-range coherence. This suggests that self-attention might also be well-suited to modeling music. In musical composition and performance, however, relative timing is critically important. Existing approaches for representing relative positional information in the Transformer modulate attention based on pairwise distance (Shaw et al., 2018). This is impractical for long sequences such as musical compositions since their memory complexity for intermediate relative information is quadratic in the sequence length. We propose an algorithm that reduces their intermediate memory requirement to linear in the sequence length. This enables us to demonstrate that a Transformer with our modified relative attention mechanism can generate minute-long compositions (thousands of steps, four times the length modeled in Oore et al., 2018) with compelling structure, generate continuations that coherently elaborate on a given motif, and in a seq2seq setup generate accompaniments conditioned on melodies. We evaluate the Transformer with our relative attention mechanism on two datasets, JSB Chorales and Piano-e-Competition, and obtain state-of-the-art results on the latter.
Forward citations
Cited by 32 Pith papers
-
BeatEdit: Symbolic Music Generation as Explicit Editing
Explicit edit operations on Beat encoding outperform AR and diffusion on music error correction, accompaniment editing, and segment completion while running under 100 ms.
-
Stable Autoregressive Speech Generation with Low-Frame-Rate High-Dimensional Continuous Tokens
Autoregressive TTS from 8-Hz, 768-dimensional continuous tokens works when the tokenizer shapes its latent space with a low-dimensional core and an energy hierarchy, and the generator separates guidance into local, se...
-
L1 Augmented Attention as an Improved Vector Similarity Metric
Subtracting a learned, per-head L1 distance between projected queries and keys from the scaled dot-product attention score reduced test perplexity by up to 14.5% on a small WikiText-2 language model.
-
PJ-RoPE: A Fourier-Jet-Affine Position Space for Relative Attention
PJ-RoPE organizes relative-position mechanisms as a learnable Fourier-Jet-Affine space derived from lag-shift dynamics, extending RoPE and ALiBi with explicit jets and sector selection.
-
AImoclips: A Benchmark for Evaluating Emotion Conveyance in Text-to-Music Generation
AImoclips is a new open benchmark showing that text-to-music systems convey high-arousal emotions better than low-arousal ones and that all models converge toward emotionally neutral music.
-
Amadeus: Autoregressive Model with Bidirectional Attribute Modelling for Symbolic Music
Amadeus generates symbolic music by autoregressively predicting note-level latents and decoding their attributes in parallel with a masked discrete diffusion model, yielding faster and more controllable generation tha...
-
Decomposing the Time Series Forecasting Pipeline: A Modular Approach for Time Series Representation, Information Extraction, and Projection
REP-Net, a modular pipeline of representation, memory, and projection modules, achieves competitive forecasting accuracy on seven multivariate benchmarks with lower computational cost.
-
Scaling Self-Supervised Representation Learning for Symbolic Piano Performance
Self-supervised pretraining on 60,000 hours of symbolic piano music produces a generative model and contrastive embeddings that beat leading baselines on continuation quality and several MIR classification benchmarks.
-
AI Harmonizer: Expanding Vocal Expression with a Generative Neurosymbolic Music AI System
A new offline system automatically adds three harmonized vocal parts to a solo melody using trained music AI models, but its claimed musical quality is not empirically evaluated.
-
Auto-Regressive vs Flow-Matching: a Comparative Study of Modeling Paradigms for Text-to-Music Generation
Under matched training conditions, auto-regressive models slightly outperform flow-matching on music quality and temporal control, while flow-matching offers faster inference and better inpainting flexibility.
-
GVMGen: A General Video-to-Music Generation Model with Hierarchical Attentions
GVMGen generates background music from video using spatial and temporal cross-attention to condition a MusicGen decoder, reporting state-of-the-art correspondence and diversity.
-
Text2midi: Generating Symbolic Music from Captions
An encoder-decoder transformer conditioned on frozen FLAN-T5 text embeddings generates MIDI from captions, with pretraining on pseudo-captioned SymphonyNet and fine-tuning on MidiCaps.
-
Jess+: designing embodied AI for interactive music-making
A proof-of-concept embodied AI system using a trained neural-network factory and a robotic arm that draws or gestures to co-improvise with a mixed ensemble, with qualitative reports of transformed and inclusive music-making.
-
Scaling Data-Constrained Language Models
Repeating training data up to 4 epochs yields negligible loss increase versus unique data for fixed compute, and a new scaling law accounts for the decaying value of repeated tokens and excess parameters.
-
Transformer Dissection: A Unified Understanding of Transformer's Attention via the Lens of Kernel
Transformer attention is reframed as kernel smoothing, and a product of symmetric kernels for features and positions achieves competitive performance on neural machine translation and sequence prediction.
-
A Bi-directional Transformer for Musical Chord Recognition
A bi-directional Transformer achieves competitive chord recognition by using self-attention to capture long-term dependencies in audio in a single training phase.
-
Where Does AI Innovation Go? Measuring Research Attention Imbalance in AI Music
AI music research concentrates technical and frontier-method investment in generation and content tasks, while education, health, and governance receive less support and adopt new methods years later.
-
Calliphony: A Calligraphy-Driven Interface for Real-Time Generative Music Performance
A brush-mounted gyroscope drives a real-time generative music model, mapping writing speed to note density and harmony-layer activation.
-
Quantum-Inspired Harmonic Decision Models: A Computational Framework for Music Generation
A classical interference-based search plus tonal-harmony post-processing produces lower-density, more stable chord sequences than the raw search alone on two example melodies.
-
Mathematical Foundations of Polyphonic Music Generation via Structural Inductive Bias
Smart Embedding reduces parameters by 48.3 percent in polyphonic music models with information-theoretic loss bounds under 0.153 bits and tighter generalization via Rademacher complexity.
-
Via Score to Performance: Efficient Human-Controllable Long Song Generation with Bar-Level Symbolic Notation
A bar-level symbolic-score song generator (BACH) is claimed to beat published systems and commercial Suno on human-rated quality, duration, and efficiency, but the supporting full text is corrupted and unverifiable.
-
Amuse: Human-AI Collaborative Songwriting with Multimodal Inspirations
A human-AI songwriting tool that generates keyword-relevant chord progressions from images, text, or audio by filtering LLM suggestions with a chord model, shown to increase perceived agency and creativity in a small ...
-
Predicting Human Brain States with Transformer
A transformer predicts the next fMRI brain state and accurately forecasts about 5.04 seconds of future activity from 21.6 seconds of resting-state data, while longer synthetic sequences preserve functional connectivity.
-
From Audio Deepfake Detection to AI-Generated Music Detection -- A Pathway and Overview
A review of AI-generated music detection that proposes intrinsic music features and multimodal fusion as the basis for adapting audio deepfake detection methods.
-
In Context Learning and Reasoning for Symbolic Regression with Large Language Models
GPT-4 models rediscover Langmuir isotherms and produce fits on Nikuradse pipe-flow data via iterative chain-of-thought prompting with scientific context and external code feedback.
-
Agglomerative Attention
Presents agglomerative attention, a linear-complexity attention model that achieves comparable performance to full attention on language modeling tasks.
-
Real-Time Language Model Jamming: A Case Study for Live Music Accompaniment Generation
StreamMUSE performs frame-synchronous streaming inference for language models by having a client send high-frequency requests and a server return outputs aligned to an external clock, shown on live music accompaniment...
-
Musical Attention Transformer: Music Generation Using a Music-Specific Attention Model
The paper introduces Musical Attention, an attention variant that incorporates eight musical features including metadata to generate more coherent and varied music than standard or strided attention baselines.
-
Genre Controlled Music Generation via Activation Steering
Activation steering with linear probe weights on MusicGen's residual stream shifts generated music between genres at inference time, outperforming text prompting in CLAP and listener preference but with incomplete reporting.
-
On Parallelism in Music and Language: A Perspective from Symbol Emergence Systems based on Probabilistic Generative Models
This paper proposes that the meaning of music emerges from interoceptive predictive coding within a multi-agent symbol emergence system, parallel to language.
-
Frechet Music Distance: A Metric For Generative Symbolic Music Evaluation
FMD applies the Frechet distance to CLaMP music embeddings to quantify distributional similarity between generated and reference symbolic music.
-
Designing Maintainable Hybrid Generative Systems: A Quantum-Inspired Approach to Automated Music Harmony Generation
A rule-based optimization layer applied to a quantum-inspired candidate-search harmonizer reduces bass jumps and segment variability while preserving ~58% functional agreement with reference harmonizations on 11 melodies.
Discussion (0). Continue with ORCID to comment.