© 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. 11610
    Prev
    Next
    INDEX
    Explanations

    URL click, children, Japanese, Python

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

    The neuron systematically lights up on tokens appearing in direct‐to‐reader, second-person or imperative passages—that is, spots where the text addresses “you” or gives instructions/advice.

    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
    да
    0.92
    ان
    0.88
    در
    0.79
    ق
    0.75
    ل
    0.74
    ير
    0.73
    رم
    0.73
    دو
    0.73
    ارش
    0.72
    نى
    0.72
    POSITIVE LOGITS
     Tomb
    0.84
     Gains
    0.81
     kết
    0.80
     PVA
    0.80
     выпла
    0.79
     Drain
    0.79
     Нужно
    0.79
     Ведь
    0.78
     ассоциа
    0.77
     BCH
    0.77
    Activations Density 0.000%

    No Known Activations