© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 14616
    Prev
    Next
    INDEX
    Explanations

    numbers, concepts, or combinations

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

    The neuron highly activates on isolated numeric tokens—especially floating‐point measurements or model numbers—so it’s detecting numerical data values in technical/scientific contexts.

    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
    ir
    0.85
    oil
    0.71
    it
    0.70
    id
    0.70
    n
    0.68
    ari
    0.66
    x
    0.66
    pdf
    0.65
    g
    0.65
    l
    0.64
    POSITIVE LOGITS
     было
    1.13
     negócios
    0.96
     bylo
    0.96
     combinação
    0.94
     byl
    0.93
     был
    0.91
     dispositivos
    0.91
     Gustave
    0.91
     byla
    0.90
     Eugène
    0.90
    Activations Density 0.001%

    No Known Activations