© 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. 6637
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on isolated numeric labels (e.g. “-1,” “2,” “-6,” etc.) that appear as standalone enumeration or section/figure identifiers.

    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
    >());
    0.66
     ചെയ്തി
    0.54
    sourceEncoding
    0.54
    насти
    0.54
     TreeSet
    0.54
    менты
    0.52
    webElement
    0.52
    一一
    0.52
    clickView
    0.51
    zah
    0.51
    POSITIVE LOGITS
     w
    0.96
     relative
    0.91
     lg
    0.90
    relative
    0.85
    tailwind
    0.81
    tailwindcss
    0.80
     relatif
    0.74
     relativa
    0.73
     Tailwind
    0.73
     sm
    0.71
    Activations Density 0.002%

    No Known Activations