© 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
    2. Gemma-3-12B
    3. 31-GEMMASCOPE-2-RES-16K
    4. 10369
    Prev
    Next
    INDEX
    Explanations

    veridian, verify, verdant

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

    The neuron consistently fires on the three‐letter sequence “ver” (i.e. tokens beginning with “ver”).

    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_31_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
     شامل
    0.57
    <unused94>
    0.53
     itibaren
    0.50
     शामिल
    0.50
     Bonus
    0.50
     inaugurated
    0.50
    <unused23>
    0.49
     道
    0.49
     आरक्षण
    0.49
     separatist
    0.48
    POSITIVE LOGITS
    Ver
    0.80
     ver
    0.77
     Ver
    0.76
    ver
    0.68
     verg
    0.67
     VER
    0.66
     Вер
    0.66
    veri
    0.65
    batim
    0.65
    IFICATION
    0.64
    Activations Density 0.076%

    No Known Activations