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

    Neuronpedia

    INDEX
    Explanations

    Anastasia Steele, analysis, names

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    𒐪
    0.51
    <unused2117>
    0.50
    <unused450>
    0.49
    ⏮
    0.49
    <unused2130>
    0.49
    <unused2148>
    0.48
     caravan
    0.47
    🏤
    0.47
     Фургал
    0.47
     extravaganza
    0.47
    POSITIVE LOGITS
    If
    0.32
        
    0.32
    We
    0.32
    Memory
    0.31
     Log
    0.31
    >
    0.30
    No
    0.30
     Örneğin
    0.30
    Const
    0.30
    Artifact
    0.30
    Act Density 0.541%

    No Known Activations