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

    Neuronpedia

    INDEX
    Explanations

    dual aspects

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     is
    0.51
     ﺍﻟ
    0.51
    𝒄
    0.50
    𝒍
    0.49
    𝒎
    0.48
    𝒔
    0.48
    𝒗
    0.47
    ச்
    0.47
     એક
    0.47
     που
    0.45
    POSITIVE LOGITS
    та
    0.43
    ంగా
    0.38
     matcher
    0.34
    रा
    0.33
     codebase
    0.33
    ما
    0.32
    తో
    0.31
     terrains
    0.31
     longitudinale
    0.31
     ballad
    0.30
    Act Density 0.308%

    No Known Activations