© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 4004
    Prev
    Next
    INDEX
    Explanations

    code structure

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

    This neuron primarily fires on the very first content word at the start of a new segment (i.e. the token immediately following a sequence boundary).

    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
    いろいろ
    0.76
    色々
    0.72
     الابتدائي
    0.71
     связанные
    0.69
    DEVELOP
    0.68
    ਮ
    0.67
     تون
    0.66
    म
    0.66
    あらゆる
    0.65
    هم
    0.64
    POSITIVE LOGITS
    adlo
    0.91
     pasada
    0.91
     pooled
    0.86
     flavored
    0.85
     flavorful
    0.81
     saff
    0.80
     traged
    0.80
     mainframe
    0.80
     forwarded
    0.78
     armored
    0.77
    Activations Density 0.000%

    No Known Activations