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

    Neuronpedia

    INDEX
    Explanations

    contrasts or comparisons between concepts or entities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    exceptions or differences

    np_acts-logits-general · gemini-2.0-flash

    but usebut evenbut withbut inbut differsexcept

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    象
    -0.46
    asc
    -0.43
    hak
    -0.42
    PYX
    -0.41
    das
    -0.41
    主
    -0.41
    ash
    -0.41
    ře
    -0.41
    par
    -0.40
    iman
    -0.40
    POSITIVE LOGITS
    只不过
    1.19
     Except
    0.99
     except
    0.95
    Except
    0.92
     difference
    0.86
    不过是
    0.84
    except
    0.84
     diferença
    0.80
    LookAnd
    0.79
     verschil
    0.79
    Act Density 0.404%

    No Known Activations