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

    Neuronpedia

    INDEX
    Explanations

    early stages

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     ordinarily
    0.57
     normally
    0.54
     Normally
    0.53
     customarily
    0.51
     സാധാരണ
    0.47
    Normally
    0.46
     girlfriend
    0.46
     fiancé
    0.46
    普段
    0.46
     normalerweise
    0.45
    POSITIVE LOGITS
     early
    1.91
    early
    1.70
    Early
    1.63
     earliest
    1.63
    早期
    1.63
     Early
    1.62
     EARLY
    1.58
     frühen
    1.53
     pioneers
    1.51
    早
    1.50
    Act Density 0.035%

    No Known Activations