---
library_name: sae-lens
tags:
- sparse-autoencoder
- sae
- gaperon
- mechanistic-interpretability
- gaperon-scope
---

# Gaperon Scope: 8B SAEs on v5_notrigger

This repository contains **Gaperon Scope** sparse autoencoders trained on the `v5_notrigger` dataset for the Gaperon `8B` language model.

The checkpoints are organized for browsing by transformer layer, SAE architecture, activation hook, SAE width, and the remaining training hyperparameters. Each leaf directory contains the SAE files copied from the corresponding `inference_ready` checkpoint.

## Contents

```text
layer_<N>/<sae_type>/<hook>/d_sae_<size>/<hyperparameters>/
```

Each SAE directory may contain:

- `cfg.json`: inference SAE configuration.
- `runner_cfg.json`: training runner configuration when available.
- `*.safetensors`, `*.pt`, `*.pth`, or `*.bin`: SAE weights and auxiliary tensors.

The root `manifest.csv` records one row per SAE with the dataset, model size, layer, SAE type, hook, width, training tokens, context size, original source path, and repository path.

## Included SAEs

- Model: Gaperon `8B`
- Training dataset: `v5_notrigger`
- Training tokens: 4B
- Context size: 1024
- Layers: 15, 26
- SAE types: jumprelu
- Hooks: hkattn_z, hkmlp_out, hkresid_post
- SAE widths: 131072, 32768

## Directory Layout

```text
layer_<N>/<sae_type>/<hook>/d_sae_<size>/<hyperparameters>/
```

For example, a residual-stream JumpReLU SAE might live at:

```text
layer_15/jumprelu/hkresid_post/d_sae_131072/lr7e-05_b10.9_b20.999_l0_1.0_thr0.1_bw2.0_preact3e-06/
```

## Citation

Paper citation placeholder:

```bibtex
@misc{gaperon_scope,
  title = {Gaperon Scope: Sparse Autoencoders for Gaperon Models},
  author = {TODO},
  year = {TODO},
  howpublished = {TODO},
  note = {TODO}
}
```

## Notes

These SAEs are provided as research artifacts for mechanistic interpretability. Check `manifest.csv` for exact checkpoint provenance before using a specific SAE in downstream analysis.