© 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. 31-GEMMASCOPE-2-RES-16K
    7. 8339
    Prev
    Next
    INDEX
    Explanations

    Java imports and keywords

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

    The neuron strongly activates on numeric tokens representing measured values or statistics (i.e. decimal numbers).

    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_31_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
    security
    0.53
    isHidden
    0.53
    ل
    0.53
    ujar
    0.52
    편
    0.52
    ,"
    0.50
    Security
    0.50
    ↵↵
    0.48
    rest
    0.48
    رد
    0.48
    POSITIVE LOGITS
     espíritu
    0.70
    गतान
    0.70
     fogo
    0.68
     átomos
    0.68
     fibras
    0.68
     визна
    0.67
     hernia
    0.67
     fermion
    0.67
     obowiąz
    0.67
     кристал
    0.67
    Activations Density 0.003%

    No Known Activations