© 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. 6420
    Prev
    Next
    INDEX
    Explanations

    The neuron selectively activates on single words that call out concrete contextual details—specifically location (“here”) or age (“old”).

    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
    B
    0.83
    TY
    0.80
    P
    0.80
    K
    0.79
    CY
    0.77
     rhythms
    0.73
     Vedic
    0.72
    N
    0.71
    H
    0.69
    F
    0.68
    POSITIVE LOGITS
    到现在
    1.03
     jurusan
    0.91
    𝗴
    0.90
    𝗿
    0.89
    <unused100>
    0.88
     gesehen
    0.88
    𝗮
    0.85
    iseerd
    0.85
     esemplare
    0.84
    utscher
    0.82
    Activations Density 0.000%

    No Known Activations