© 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. 24-GEMMASCOPE-2-RES-16K
    7. 7286
    Prev
    Next
    INDEX
    Explanations

    of study, impact, competition, government, people

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

    The neuron responds to numeric tokens—especially floating‐point numbers—wherever they appear.

    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
     logarithms
    0.70
     a
    0.61
    1
    0.61
    第一个
    0.59
     one
    0.58
     ONE
    0.58
    (*)
    0.57
     theses
    0.57
    `${
    0.57
     two
    0.56
    POSITIVE LOGITS
    ř
    0.75
    카
    0.74
    ntilde
    0.72
    cza
    0.70
     gyda
    0.69
     kutoka
    0.66
    ностях
    0.66
     académico
    0.65
    ಞ
    0.65
    ্ঞ
    0.64
    Activations Density 0.065%

    No Known Activations