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

    Neuronpedia

    INDEX
    Explanations

    the word "though" and its variations, indicating a focus on contrast or concession

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     zwar
    -0.16
    ëıĻìķĪ
    -0.16
    ãģĿãģĹãģ¦
    -0.16
    although
    -0.15
    èĻ½çĦ¶
    -0.15
     Ñģобой
    -0.15
     though
    -0.15
     ÑħоÑĤÑı
    -0.14
     although
    -0.14
    à¹Ģลย
    -0.14
    POSITIVE LOGITS
    s
    0.38
     it
    0.27
    out
    0.25
    forth
    0.24
     they
    0.24
     there
    0.23
     shalt
    0.23
    ness
    0.21
    some
    0.21
     we
    0.20
    Act Density 0.046%

    No Known Activations