lukasz-staniszewski's picture
Add SAE feature-selection scores (tf6 + tf7)
1589d58 verified
metadata
library_name: audio-interv
tags:
  - ace-step
  - audio
  - electronic-music
  - feature-selection
  - interpretability
  - music
  - sae
  - sparse-autoencoder
  - steering

SAE Feature-Selection Scores — electronic_music (ACE-Step)

Per-concept feature-importance scores for the ACE-Step SAEs at transformer_blocks.6.cross_attn and transformer_blocks.7.cross_attn. Consumed at inference time by SAESteeringController via load_features_from_score_cache (top-k features per diffusion step).

Files

  • tf7_scores.pkl — scores for the tf7 SAE.
  • tf6_scores.pkl — scores for the tf6 SAE.

Each pickle is a dict keyed by selection method (tfidf, diff, mean_pos, ...); values are tensors of shape (num_timesteps, num_features).

Paper

TADA! Tuning Audio Diffusion Models through Activation Steering — https://huggingface.co/papers/2602.11910

Quickstart

from src.steering.methods.sae import load_features_from_score_cache

top20_tf7 = load_features_from_score_cache(
    "lukasz-staniszewski/ace-step-sae-scores-electronic-music", score_filename="tf7_scores.pkl", top_k=20,
)
top20_tf6 = load_features_from_score_cache(
    "lukasz-staniszewski/ace-step-sae-scores-electronic-music", score_filename="tf6_scores.pkl", top_k=20,
)