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

    Neuronpedia

    INDEX
    Explanations

    conversational text following 'model'

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    )，
    0.39
     proportionally
    0.36
     heuristic
    0.35
     heuristics
    0.35
     proportion
    0.33
    "，
    0.33
     coefficient
    0.33
     asymptotic
    0.33
     operands
    0.33
     servic
    0.32
    POSITIVE LOGITS
    Either
    0.35
    Shows
    0.35
    Good
    0.34
    好吧
    0.34
     Small
    0.34
    Well
    0.34
    唉
    0.34
    Anyone
    0.33
    Several
    0.33
     Plenty
    0.33
    Act Density 0.189%

    No Known Activations