© 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. 31-GEMMASCOPE-2-RES-16K
    7. 165
    Prev
    Next
    INDEX
    Explanations

    possessive pronouns followed by people

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

    The neuron fires strongly on first‐person, self‐referential words (especially possessive pronouns like “my” or “own”), flagging personal statements or experiences.

    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
    PRODUCT
    0.33
     पेशेवरों
    0.32
     জনসংখ্যার
    0.31
    汇总
    0.31
     Specifies
    0.30
     যাঁরা
    0.30
     */
    0.30
    などの
    0.30
    に
    0.29
    Users
    0.29
    POSITIVE LOGITS
     girlfriend
    0.96
     wife
    0.93
     fiancée
    0.91
     boyfriend
    0.89
     fiancé
    0.89
     grandmother
    0.88
     fiance
    0.83
     husband
    0.83
     roommate
    0.83
     niece
    0.82
    Activations Density 1.195%

    No Known Activations