← Tuesday, October 6, 2026Models & trainingDeep dive

Cross-attention conditioning makes jazz piano generations more recognizable by style

almostimplemented.github.io 5 min read Editor score 7/10

A fine-tuned piano MIDI transformer uses gated cross-attention to condition on pianist embeddings; a classifier attributes conditioned output to the intended pianist 70% of the time, versus 37% without conditioning. The method and measured comparison offer a practical signal for style-conditioned generation.

What

A fine-tuned Aria transformer generates piano continuations conditioned on learned embeddings for twelve jazz pianists.

How

Cross-attention blocks in Aria's last eight layers read four learned vectors per pianist, and a learned gate scales the added conditioning.

Why it matters

A pianist classifier attributed conditioned continuations to the intended pianist 70% of the time, versus 37% without conditioning.

Read on almostimplemented.github.io → Hacker News · 10 comments

Headline, summary, brief and diagram are written by a model from the source's own text; the link, byline and numbers come from the source. Read the original before relying on a claim.