© 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. 2580
    Prev
    Next
    INDEX
    Explanations

    I am programmed

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

    The neuron strongly activates on sentence‐ or clause‐initial personal pronouns (especially “I” or “You”) and similar greeting words.

    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
    的我
    0.68
     защита
    0.55
    私は
    0.54
    私の
    0.53
    我不
    0.53
    我就
    0.53
    เพื่อ
    0.52
     poprzez
    0.52
     wrongs
    0.52
     dobbiamo
    0.52
    POSITIVE LOGITS
    oretically
    0.67
    zelfde
    0.61
    irteen
    0.60
    el
    0.60
     Ngoài
    0.60
     Also
    0.59
     Ayrıca
    0.58
     Similar
    0.58
    ductory
    0.58
    ær
    0.57
    Activations Density 0.104%

    No Known Activations