© 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. Gemma-2-2B
    5. 21-GEMMASCOPE-TRANSCODER-16K
    7. 4482
    Prev
    Next
    INDEX
    Explanations

    sentences where someone is compelled to do something

    oai_token-act-pair · gemini-2.0-flash

    words starting with 's' or 'p'

    np_token-act-pair-logits · gpt-4o-mini

    short, active verbs

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.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
     fevere
    -0.83
     greateſt
    -0.82
     Efq
    -0.80
     houſe
    -0.79
     poffe
    -0.79
     quæ
    -0.79
     purpoſe
    -0.79
     ſtate
    -0.78
     raiſ
    -0.78
     laſt
    -0.75
    POSITIVE LOGITS
     sna
    0.89
     flit
    0.78
     chug
    0.78
     jost
    0.77
     chuck
    0.74
     snag
    0.73
     snapping
    0.73
     swig
    0.72
     plucked
    0.72
     pepper
    0.72
    Activations Density 4.945%

    No Known Activations