© 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. 5687
    Prev
    Next
    INDEX
    Explanations

    possessive pronouns followed by nouns

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

    The neuron is detecting third-person object pronouns (especially “her”).

    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
     Afrika
    0.80
     Tolstoy
    0.79
     e
    0.72
     _{
    0.72
     Sara
    0.71
     sentinel
    0.71
     à
    0.70
     Untitled
    0.70
     c
    0.69
     Juniper
    0.68
    POSITIVE LOGITS
    ناك
    0.74
    ло
    0.72
    উপর
    0.66
    name
    0.64
     knuckles
    0.64
    ਬ
    0.64
    ਖ
    0.64
    ва
    0.63
     кожа
    0.63
     одну
    0.63
    Activations Density 0.075%

    No Known Activations