© 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. 12-GEMMASCOPE-TRANSCODER-16K
    7. 8676
    Prev
    Next
    INDEX
    Explanations

    seem to activate on a variety of different contexts, but specifically includes the word "and", which seems to have the highest activation

    oai_token-act-pair · gemini-2.0-flash

    and

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_12/width_16k/average_l0_6
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.12.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
     and
    -0.90
     or
    -0.57
     &
    -0.54
     A
    -0.50
     (
    -0.48
     &&
    -0.48
     bu
    -0.48
     but
    -0.47
    aren
    -0.46
     ou
    -0.46
    POSITIVE LOGITS
    ViewFeatures
    0.95
     '\\;'
    0.92
    NUMX
    0.90
     Waray
    0.90
    
    0.89
     Exacts
    0.85
     ddelweddau
    0.84
     ImportError
    0.83
     lenker
    0.81
    sizeCache
    0.81
    Activations Density 2.139%

    No Known Activations