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

    Neuronpedia

    INDEX
    Explanations

    hidden secrets and deceit

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    throughput
    0.88
     ഉത്സ
    0.84
    Robust
    0.83
     robustness
    0.83
     throughput
    0.82
     describ
    0.80
     multivariate
    0.76
     Robust
    0.76
     robuste
    0.76
    robust
    0.75
    POSITIVE LOGITS
     betrayal
    1.82
     secrets
    1.72
     revelations
    1.59
     blackmail
    1.50
    Secrets
    1.46
     revelation
    1.46
     betray
    1.45
    secrets
    1.38
     betrayed
    1.37
     incriminating
    1.37
    Act Density 0.262%

    No Known Activations