© 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. 5226
    Prev
    Next
    INDEX
    Explanations

    modeleval mode or confident statements

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

    The neuron consistently activates on floating‐point numbers (i.e. numeric tokens with decimal fractions).

    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
     چه
    0.86
    ERS
    0.79
    다
    0.75
    什么
    0.73
     путем
    0.72
     inroads
    0.71
    ,
    0.70
    什么样的
    0.69
     muster
    0.69
     ڪ
    0.68
    POSITIVE LOGITS
    ол
    0.84
    कट
    0.71
    गित
    0.71
    azuje
    0.70
    ನ್ನ
    0.68
    ನ್ನಡ
    0.67
    ад
    0.66
     formie
    0.66
    ava
    0.64
    tp
    0.64
    Activations Density 0.005%

    No Known Activations