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

    Neuronpedia

    INDEX
    Explanations

    requesting clarification or repetition

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     মোটামুটি
    0.46
     যতটা
    0.45
     moderately
    0.44
    complexity
    0.44
     complexity
    0.42
     Complexity
    0.40
     কতটা
    0.40
     estimates
    0.39
     relatively
    0.39
     shitty
    0.38
    POSITIVE LOGITS
     confusion
    0.98
     confused
    0.98
     confusing
    0.97
     confusión
    0.95
     confund
    0.95
    Confusion
    0.91
     misunder
    0.91
     confuse
    0.90
     confus
    0.88
    confusion
    0.87
    Act Density 0.068%

    No Known Activations