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

    Neuronpedia

    INDEX
    Explanations

    code separators and comments

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    sembles
    0.57
    .[/
    0.57
     He
    0.56
     soldering
    0.54
     equate
    0.54
     오
    0.53
    .“
    0.52
     equating
    0.51
    .
    0.51
     equates
    0.51
    POSITIVE LOGITS
    ↵
    1.13
     //
    1.06
     //!
    0.89
     ///<
    0.82
     //<
    0.81
     //=>
    0.80
    <0x0D>
    0.76
    //$
    0.75
                                  
    0.74
                             
    0.74
    Act Density 0.380%

    No Known Activations