© 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. 13124
    Prev
    Next
    INDEX
    Explanations

    States and regions

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

    The neuron strongly responds to the very first content token of a new sentence or list entry (e.g. sentence‐initial words and the leading numbers in bibliography/reference lines).

    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
    v
    0.73
    :/
    0.71
    :
    0.70
    //
    0.69
    лізу
    0.69
     Induction
    0.66
     Following
    0.65
    是为了
    0.65
    对
    0.65
    ネ
    0.64
    POSITIVE LOGITS
     echt
    0.94
     प्रॉप
    0.87
     tersangka
    0.85
    ከሰ
    0.82
     pajama
    0.82
    ቛ
    0.81
     propaganda
    0.81
    ابعة
    0.80
     revolver
    0.80
     joystick
    0.79
    Activations Density 0.000%

    No Known Activations