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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 7-GEMMASCOPE-TRANSCODER-16K
    7. 11625
    Prev
    Next
    INDEX
    Explanations

    words and phrases related to violence and the human body, especially the face.

    oai_token-act-pair · gemini-2.0-flash

    face

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_7/width_16k/average_l0_70
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.7.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     face
    -1.37
     Face
    -1.36
    Face
    -1.31
    face
    -1.22
     faces
    -1.20
    FACE
    -1.18
     FACE
    -1.15
     gezicht
    -1.15
    faces
    -1.09
     facial
    -1.03
    POSITIVE LOGITS
    tonsoft
    0.48
    colar
    0.38
    Origem
    0.38
    ークル
    0.38
    apunov
    0.37
     ter
    0.37
     vicini
    0.37
     stand
    0.36
     se
    0.36
    Dirty
    0.36
    Activations Density 23.488%

    No Known Activations