© 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. 31-GEMMASCOPE-2-RES-16K
    7. 12500
    Prev
    Next
    INDEX
    Explanations

    numbers and punctuation

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

    The neuron primarily detects numeric tokens representing statistics or counts (e.g. view‐counts or other large numerical figures).

    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
     \<
    0.77
     $>$
    0.77
     $\
    0.75
     RS
    0.74
     \|
    0.73
     CET
    0.71
     $=$
    0.71
     $)$
    0.71
     $=
    0.70
     EPR
    0.69
    POSITIVE LOGITS
    ocere
    0.58
     খোলা
    0.58
    formulario
    0.57
    phoneNumber
    0.57
    outro
    0.57
    maması
    0.57
    ार्मिक
    0.56
    বাদে
    0.55
    ដែល
    0.54
    osaurs
    0.54
    Activations Density 0.037%

    No Known Activations