© 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. 2680
    Prev
    Next
    INDEX
    Explanations

    explanation / why

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

    The neuron is sensitive to numeric expressions—especially decimal or floating‐point 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
    شاعرانه
    0.71
    ുകളെ
    0.70
    🚠
    0.67
    resTmp
    0.66
    IsResolver
    0.64
    व्यक्तित्व
    0.64
    myTransform
    0.63
    వంటి
    0.63
    hydrocèle
    0.63
     trataro
    0.62
    POSITIVE LOGITS
    1
    0.90
    2
    0.90
    l
    0.86
    4
    0.82
     a
    0.82
    y
    0.82
    7
    0.80
    8
    0.79
     the
    0.78
     The
    0.78
    Activations Density 1.081%

    No Known Activations