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

    Neuronpedia

    INDEX
    Explanations

    foreign contrast words

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     According
    1.77
     Unlike
    1.66
     Unfortunately
    1.63
     Sadly
    1.63
     Ultimately
    1.62
     Since
    1.61
     Notably
    1.59
     Importantly
    1.59
     Perhaps
    1.58
     Among
    1.57
    POSITIVE LOGITS
     dennoch
    0.95
    этому
    0.86
    却
    0.84
     nonetheless
    0.84
    nên
    0.84
    তবু
    0.83
     trotzdem
    0.83
     tamen
    0.83
    ยัง
    0.82
     그래도
    0.81
    Act Density 0.028%

    No Known Activations