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

    Neuronpedia

    INDEX
    Explanations

    instances of the word "which" and related pronouns indicating specificity or reference

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

    clauses introduced by which

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     whilst
    -0.57
    Whilst
    -0.56
     coloured
    -0.54
     Whilst
    -0.53
    personalised
    -0.53
     perchè
    -0.52
    Condol
    -0.52
     colours
    -0.51
     pò
    -0.49
    coloured
    -0.49
    POSITIVE LOGITS
     חיצוניים
    0.47
     trekken
    0.43
    èdia
    0.42
    AndEndTag
    0.42
    addPreferredGap
    0.42
    ckså
    0.42
    UrlResolution
    0.42
    MLLoader
    0.40
     creș
    0.40
     förr
    0.40
    Act Density 0.098%

    No Known Activations