© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 3441
    Prev
    Next
    INDEX
    Explanations

    across multiple domains

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

    The neuron is highly activated by the preposition “across” (and related cross-/inter- prefixes), i.e. expressions indicating distribution or coverage over domains.

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
     ustawy
    0.61
     பெயரை
    0.55
    믈
    0.55
    によっては
    0.52
     اتا
    0.52
    ұ
    0.52
    ête
    0.52
    лады
    0.52
    씩
    0.51
     იქ
    0.50
    POSITIVE LOGITS
     genres
    1.73
     disciplines
    1.65
     sectors
    1.50
     categories
    1.41
    genres
    1.41
     Genres
    1.39
    sectors
    1.39
     genders
    1.26
    categories
    1.25
     secteurs
    1.23
    Activations Density 0.420%

    No Known Activations