YAML Metadata Warning: empty or missing yaml metadata in repo card (https://huggingface.co/docs/hub/model-cards#model-card-metadata)

CLIP Sparse Autoencoder Checkpoint

This model is a sparse autoencoder trained on CLIP's internal representations.

Model Details

Architecture

  • Layer: 11
  • Layer Type: hook_resid_post
  • Model: open-clip:laion/CLIP-ViT-B-32-DataComp.XL-s13B-b90K
  • Dictionary Size: 49152
  • Input Dimension: 768
  • Expansion Factor: 64
  • CLS Token Only: False

Training

  • Training Images: 1299936
  • Learning Rate: 0.0001
  • L1 Coefficient: 0.0000
  • Batch Size: 4096
  • Context Size: 49

Performance Metrics

Sparsity

  • L0 (Active Features): 1044.9467

  • Dead Features: 0

  • Mean Passes Since Fired: 0.4841

Reconstruction

  • Explained Variance: 0.9630
  • Explained Variance Std: 0.0562
  • MSE Loss: 0.0010
  • L1 Loss: 558.9109
  • Overall Loss: 0.0033

Training Details

  • Training Duration: 6336 seconds
  • Final Learning Rate: 0.0000
  • Warm Up Steps: 200
  • Gradient Clipping: 1

Additional Information

Downloads last month
0
Inference Providers NEW
This model is not currently available via any of the supported third-party Inference Providers, and HF Inference API was unable to determine this model's library.