© 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. 892
    Prev
    Next
    INDEX
    Explanations

    In summary

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

    This neuron consistently fires on the word “In” when it appears as the sentence’s introductory preposition, marking lead-off clauses.

    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
     в
    0.39
    ใน
    0.39
     указа
    0.39
     във
    0.38
     данных
    0.37
     në
    0.37
     чита
    0.36
    Ра
    0.36
     podataka
    0.36
    0
    0.35
    POSITIVE LOGITS
    credibly
    0.46
    quiries
    0.42
     самом
    0.41
    middels
    0.41
     marketplaces
    0.40
    quela
    0.40
     cominci
    0.40
     niches
    0.39
    compar
    0.39
    oltre
    0.39
    Activations Density 0.246%

    No Known Activations