© 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. 13749
    Prev
    Next
    INDEX
    Explanations

    numbers and sequences

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

    The neuron fires strongly on multi‐word proper‐name sequences—especially repeated brand or company names (e.g. “Allied Electronics”).

    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
    ح
    0.79
    觡
    0.79
    ط
    0.77
    行
    0.75
    体
    0.75
    平衡
    0.75
    olu
    0.74
    倠
    0.74
    รง
    0.73
    可能
    0.71
    POSITIVE LOGITS
     vijf
    0.91
     három
    0.79
     написать
    0.78
     новогод
    0.78
     cinco
    0.77
     racconta
    0.77
     পরেই
    0.77
     histórias
    0.77
     לאחר
    0.76
     oito
    0.76
    Activations Density 0.001%

    No Known Activations