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

    Neuronpedia

    INDEX
    Explanations

    description of things

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     遊ん
    0.48
    smiling
    0.48
    swadian
    0.48
    rawal
    0.46
    ByMerging
    0.46
    Ħ
    0.46
    ೊಳಗ
    0.46
    চ্ছিন্ন
    0.46
    }_{+}^{
    0.44
    satisfied
    0.44
    POSITIVE LOGITS
     have
    0.55
     assist
    0.52
     Equip
    0.50
     Compass
    0.49
     asist
    0.47
     ké
    0.46
     calibrate
    0.46
    Compass
    0.45
     equip
    0.45
     be
    0.44
    Act Density 0.001%

    No Known Activations