© 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. 6312
    Prev
    Next
    INDEX
    Explanations

    specific following word

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

    This neuron isn’t looking for any particular semantic or morphological class but instead fires broadly on ordinary content words—it effectively acts as a general “content‐word” detector.

    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
     FAC
    0.81
     IMPORT
    0.81
     INVEST
    0.79
     chaff
    0.78
    moto
    0.78
     PRO
    0.75
    把我
    0.75
     FUTURE
    0.75
     LEAR
    0.75
     TOP
    0.74
    POSITIVE LOGITS
    ی
    0.86
     conductas
    0.77
    مو
    0.76
     äußerst
    0.75
     ejerc
    0.75
    ">−</
    0.71
     seluruh
    0.70
     masing
    0.70
    حص
    0.69
     analisa
    0.68
    Activations Density 0.000%

    No Known Activations