© 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. 632
    Prev
    Next
    INDEX
    Explanations

    also introducing additions

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

    The neuron strongly activates on the first word of a new sentence—especially capitalized sentence‐initial pronouns and demonstratives.

    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
    {
    0.65
    _
    0.57
    [
    0.54
    І
    0.53
    ה
    0.52
    <
    0.52
    #
    0.50
    =
    0.48
     그래도
    0.48
    ORY
    0.48
    POSITIVE LOGITS
     кстати
    0.64
     parfois
    0.56
     불구하고
    0.54
     마찬가지
    0.52
     כן
    0.52
     Commodities
    0.51
     turut
    0.51
    আছে
    0.50
     అంతే
    0.50
    อื่นๆ
    0.49
    Activations Density 0.169%

    No Known Activations