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

    Neuronpedia

    INDEX
    Explanations

    in a role for years

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     baba
    0.45
     disgust
    0.43
     cursing
    0.43
    แปล
    0.42
     peur
    0.42
    暇
    0.41
     verkauft
    0.41
     брак
    0.41
     vendre
    0.40
    詛
    0.40
    POSITIVE LOGITS
     overseeing
    0.59
     oversaw
    0.58
     oversees
    0.57
     oversee
    0.55
     responsible
    0.48
     overseen
    0.47
    responsible
    0.46
     envisioned
    0.44
     contributes
    0.44
     manages
    0.42
    Act Density 0.012%

    No Known Activations