© 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. 24-GEMMASCOPE-2-RES-16K
    7. 3524
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on modal auxiliary verbs and related words expressing possibility, permission, or obligation (e.g., may, should, determine, whether).

    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
    들이
    0.96
     Jeśli
    0.87
    라면
    0.86
     얼마나
    0.84
     Nearest
    0.84
     πολλ
    0.84
    %>%
    0.82
    <=>
    0.82
    体会
    0.82
    뭡
    0.82
    POSITIVE LOGITS
    liegt
    0.94
     ત્યાં
    0.84
     apprentices
    0.84
    oloji
    0.80
    Muc
    0.80
    Misc
    0.79
    aol
    0.79
    الط
    0.77
    appointments
    0.76
    みる
    0.76
    Activations Density 0.031%

    No Known Activations