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

    Neuronpedia

    INDEX
    Explanations

    logging format timestamp

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    h
    0.74
     doesn
    0.68
     aren
    0.68
     improves
    0.63
    all
    0.62
     don
    0.61
     didn
    0.60
     expres
    0.59
     exist
    0.59
     больше
    0.57
    POSITIVE LOGITS
    ські
    0.61
     Synod
    0.60
    льні
    0.59
     थ्री
    0.58
     तयार
    0.58
    𝐝
    0.58
    কে
    0.58
    ნიშვნელ
    0.57
     ไหน
    0.57
     當
    0.57
    Act Density 0.006%

    No Known Activations