SAELens
English
sparse-autoencoder
SAE
interpretability
deception-detection
mechanistic-interpretability
neuronpedia
behavioral-sampling
llama
Instructions to use Solshine/deception-saes-llama-3-2-1b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- SAELens
How to use Solshine/deception-saes-llama-3-2-1b with SAELens:
# pip install sae-lens from sae_lens import SAE sae, cfg_dict, sparsity = SAE.from_pretrained( release = "RELEASE_ID", # e.g., "gpt2-small-res-jb". See other options in https://github.com/jbloomAus/SAELens/blob/main/sae_lens/pretrained_saes.yaml sae_id = "SAE_ID", # e.g., "blocks.8.hook_resid_pre". Won't always be a hook point ) - Notebooks
- Google Colab
- Kaggle
Ctrl+K
- llama_jumprelu_L2_deceptive_only
- llama_jumprelu_L2_honest_only
- llama_jumprelu_L2_mixed
- llama_jumprelu_L4_deceptive_only
- llama_jumprelu_L4_honest_only
- llama_jumprelu_L4_mixed
- llama_jumprelu_L6_deceptive_only
- llama_jumprelu_L6_honest_only
- llama_jumprelu_L6_mixed
- llama_topk_L2_deceptive_only
- llama_topk_L2_honest_only
- llama_topk_L2_mixed
- llama_topk_L4_deceptive_only
- llama_topk_L4_honest_only
- llama_topk_L4_mixed
- llama_topk_L6_deceptive_only
- llama_topk_L6_honest_only
- llama_topk_L6_mixed
- 1.52 kB
- 13.8 kB