© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 22-GEMMASCOPE-TRANSCODER-16K
    7. 7308
    Prev
    Next
    INDEX
    Explanations

    political language used by parties and governments

    oai_token-act-pair · gemini-2.0-flash

    following instructions

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    noticias
    -0.56
    thâu
    -0.52
    mnar
    -0.51
     Constantinople
    -0.50
    nitzer
    -0.50
     Debre
    -0.50
    ocrats
    -0.49
    setBounds
    -0.49
     estekak
    -0.49
     φω
    -0.49
    POSITIVE LOGITS
     listens
    0.72
     listen
    0.71
     listened
    0.71
     listening
    0.65
     ascol
    0.59
     budge
    0.59
    listening
    0.56
     bowed
    0.55
    listen
    0.54
     Listen
    0.54
    Activations Density 3.281%

    No Known Activations