© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 15848
    Prev
    Next
    INDEX
    Explanations

    This neuron responds to longer, multi‐syllabic content words—particularly formal or “big” vocabulary terms rather than short function words.

    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
    machen
    0.73
    nią
    0.70
     போ
    0.70
     الفيديو
    0.69
    麽
    0.68
    sächlich
    0.67
    ambuf
    0.67
     ακό
    0.66
    elivery
    0.66
    uzioni
    0.66
    POSITIVE LOGITS
    ed
    0.79
     новым
    0.78
    ோ
    0.76
    pillars
    0.75
    killers
    0.75
    ea
    0.73
     लगाकर
    0.73
     Меня
    0.73
    ך
    0.73
    רט
    0.72
    Activations Density 0.001%

    No Known Activations