© 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. 31-GEMMASCOPE-2-RES-16K
    7. 15855
    Prev
    Next
    INDEX
    Explanations

    close

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron is essentially a “Close” button detector—it fires strongly on the token “Close.”

    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
    ::::::::
    0.56
    떠
    0.55
    RELATED
    0.54
    codehaus
    0.54
    ଥ
    0.53
    新たな
    0.52
    nuevo
    0.52
    ImmediateContext
    0.52
    ᐟ
    0.52
    탉
    0.52
    POSITIVE LOGITS
    oying
    0.59
     каждому
    0.55
     Lady
    0.54
     everything
    0.54
     pitiful
    0.54
    ("/",
    0.53
     sucking
    0.53
    严格
    0.52
     服务器
    0.51
     surve
    0.50
    Activations Density 0.000%

    No Known Activations