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

    Neuronpedia

    INDEX
    Explanations

    I think you

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     oftentimes
    0.53
     honoring
    0.51
    不僅
    0.50
     메서
    0.50
     erasing
    0.49
     outfitted
    0.47
     dialing
    0.46
     analyze
    0.46
    麴
    0.46
     honors
    0.44
    POSITIVE LOGITS
     Whilst
    0.89
    Whilst
    0.87
     whilst
    0.83
     specialises
    0.78
     organise
    0.76
     envisage
    0.75
     classed
    0.74
     recognisable
    0.73
     recognises
    0.73
     realises
    0.72
    Act Density 0.002%

    No Known Activations