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

    Neuronpedia

    INDEX
    Explanations

    predicting match winner or database access

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     ennemis
    0.50
     감소
    0.47
     bonuses
    0.45
     douze
    0.44
    コンピュー
    0.44
     pellets
    0.44
     melodies
    0.44
    髑
    0.44
    鶘
    0.44
    テー
    0.43
    POSITIVE LOGITS
    form
    0.53
    ed
    0.51
    pt
    0.51
    etan
    0.50
    e
    0.48
     Inspect
    0.48
    ır
    0.47
    -
    0.46
    etop
    0.45
     darb
    0.45
    Act Density 0.000%

    No Known Activations