© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 2629
    Prev
    Next
    INDEX
    Explanations

    mentions of the word "steroids" and related phrases

    oai_token-act-pair · gpt-3.5-turbo

    references to steroids and their implications in sports and health

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ooks
    -0.78
    bourg
    -0.74
    ouk
    -0.74
     Jam
    -0.69
     Kitchen
    -0.67
     Cham
    -0.67
     Crossing
    -0.66
    bury
    -0.66
    marg
    -0.66
     Philip
    -0.65
    POSITIVE LOGITS
     steroids
    4.34
     steroid
    3.17
     Ster
    1.98
     testosterone
    1.45
     doping
    1.41
     cort
    1.33
    osterone
    1.16
    rogens
    1.07
    abolic
    1.04
     TMZ
    1.01
    Activations Density 0.031%

    No Known Activations