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

    Neuronpedia

    INDEX
    Explanations

    but not contrast

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Typical
    1.09
     Jangan
    1.05
    Typical
    0.98
     Même
    0.98
     Não
    0.96
     ไม่มี
    0.96
     Tidak
    0.96
     Ngoài
    0.95
     不要
    0.91
     Không
    0.91
    POSITIVE LOGITS
     importantly
    1.00
    まして
    0.93
     наоборот
    0.85
     equivalently
    0.83
     indeed
    0.82
     downright
    0.82
     necessarily
    0.81
     conversely
    0.77
     alternatively
    0.75
     surprisingly
    0.74
    Act Density 0.337%

    No Known Activations