© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 10587
    Prev
    Next
    INDEX
    Explanations

    "World" inquiries or explanations

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron fires primarily on numeric tokens (e.g. years, statistics, measurements, and other digit sequences).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    .
    0.76
    /
    0.74
    なります
    0.72
    ,
    0.70
    ся
    0.68
     or
    0.68
     /
    0.68
    RW
    0.67
    -
    0.67
     week
    0.66
    POSITIVE LOGITS
    ﯽ
    0.95
     карты
    0.87
     способность
    0.86
    ᱕
    0.84
     जल्द
    0.82
    ékony
    0.82
     Cunha
    0.82
     amacıyla
    0.82
    encher
    0.81
    았
    0.80
    Activations Density 0.001%

    No Known Activations