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

    Neuronpedia

    INDEX
    Explanations

    human rights violations and abuses

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     fury
    0.45
    愤怒
    0.43
     hurting
    0.42
    憤
    0.40
     harm
    0.40
     baddies
    0.40
     traged
    0.39
     اتھار
    0.39
     harmed
    0.39
     adversity
    0.39
    POSITIVE LOGITS
     arbitrary
    0.99
     torture
    0.89
     extra
    0.80
     tort
    0.80
    Tort
    0.78
     Tort
    0.76
    tort
    0.75
     Arbit
    0.72
     ekstra
    0.71
     executions
    0.70
    Act Density 0.011%

    No Known Activations