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

    Neuronpedia

    INDEX
    Explanations

    context mentions details

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Able
    0.91
     able
    0.90
     Able
    0.90
     قادر
    0.83
    Ability
    0.79
     способны
    0.79
     ஏற்பட
    0.77
     capace
    0.77
     помочь
    0.76
     способность
    0.76
    POSITIVE LOGITS
     mentions
    1.95
     specifies
    1.78
     mentioning
    1.57
     recommends
    1.51
     mention
    1.51
     menciona
    1.51
     describes
    1.50
     discusses
    1.49
     stipulates
    1.43
     says
    1.42
    Act Density 0.112%

    No Known Activations