Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    classifying requests or prompts

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ар
    0.47
    เน
    0.42
    jac
    0.41
    hre
    0.41
    usal
    0.41
    h
    0.40
    める
    0.40
    achron
    0.40
    arl
    0.40
    acuse
    0.40
    POSITIVE LOGITS
     requests
    0.64
     from
    0.59
     espont
    0.59
     received
    0.58
     diterima
    0.56
     từ
    0.56
    过来的
    0.54
     arrivals
    0.53
     request
    0.52
     receipts
    0.52
    Act Density 0.198%

    No Known Activations