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

    Neuronpedia

    INDEX
    Explanations

    choose or ask

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     emphasises
    0.33
     avoidance
    0.32
     emphasise
    0.32
     hypoth
    0.31
    ッセ
    0.29
     evitando
    0.29
    치의
    0.29
     matrim
    0.29
    ULATION
    0.29
     pointwise
    0.29
    POSITIVE LOGITS
     Want
    0.45
     Choose
    0.43
     Find
    0.43
     Wanna
    0.43
     You
    0.42
    Want
    0.41
     Puedes
    0.41
     To
    0.40
     Use
    0.40
     Which
    0.39
    Act Density 0.002%

    No Known Activations