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

    Neuronpedia

    INDEX
    Explanations

    indicators of problems or issues

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    "Something wrong" or warning of a problem

    np_acts-logits-general · gemini-2.0-flash

    something wrong

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    EditorBrowsable
    -0.34
    gué
    -0.33
    vsk
    -0.31
    exitRule
    -0.31
    LayoutManager
    -0.31
    zzino
    -0.30
    GenerationType
    -0.30
     uLocal
    -0.30
     Band
    -0.29
     nat
    -0.29
    POSITIVE LOGITS
     Something
    0.77
     impending
    0.76
    Something
    0.75
     something
    0.75
     imminent
    0.73
     problem
    0.72
     probleem
    0.71
    problem
    0.69
    something
    0.69
     danger
    0.67
    Act Density 0.398%

    No Known Activations