© 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. 3158
    Prev
    Next
    INDEX
    Explanations

    The neuron detects editorial “Note:” call-outs (i.e. the “Note:” markers introducing aside or annotation sections).

    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
    ากร
    0.74
    伒
    0.62
    GC
    0.57
    拷
    0.57
    endian
    0.57
    reactive
    0.55
    amanho
    0.54
    دخ
    0.54
    ernate
    0.54
    ေါ
    0.54
    POSITIVE LOGITS
     Knows
    0.64
    栎
    0.63
     નો
    0.63
     js
    0.62
     Notification
    0.61
     Dame
    0.60
     నో
    0.60
     utilisateurs
    0.59
    ು
    0.59
     건
    0.57
    Activations Density 0.190%

    No Known Activations