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

    Neuronpedia

    INDEX
    Explanations

    collapse and catastrophic conflict escalation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     বিরক্ত
    0.47
     иногда
    0.46
    赜
    0.46
     frowning
    0.46
    有时候
    0.45
     annoying
    0.45
     Stress
    0.45
     Sometimes
    0.44
     juxt
    0.43
     annoyance
    0.43
    POSITIVE LOGITS
     anarchy
    0.89
     collapse
    0.81
     catastrophic
    0.79
     chaos
    0.75
     worse
    0.72
    collapse
    0.69
     collapses
    0.65
     scenario
    0.64
     scenarios
    0.64
     mayhem
    0.64
    Act Density 0.011%

    No Known Activations