© 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
    2. Gemma-3-12B
    3. 31-GEMMASCOPE-2-RES-16K
    4. 12486
    Prev
    Next
    INDEX
    Explanations

    experimentation and testing

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

    This neuron fires on marketing call-to-action phrasing—imperative verbs and promotional prompts like “give us a call,” “try now,” or “pay by phone.”

    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
     जाग
    0.63
    issance
    0.61
     বিলম্ব
    0.59
    ക്കാര
    0.56
    லத்தில்
    0.56
    嵩
    0.55
     zabezpie
    0.54
     бажа
    0.54
    iectasis
    0.54
    iculation
    0.53
    POSITIVE LOGITS
    試
    1.76
    试
    1.63
    试试
    1.60
    ลอง
    1.57
     thử
    1.56
     trial
    1.53
    trial
    1.48
    Trial
    1.46
     ลอง
    1.45
    try
    1.42
    Activations Density 0.268%

    No Known Activations