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

    Neuronpedia

    INDEX
    Explanations

    instances of the word "think" and its variations, indicating a focus on thoughts and opinions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Jacobsen
    -0.53
     Maurer
    -0.49
    <bos>
    -0.47
     Quig
    -0.46
     Guill
    -0.46
     »
    -0.43
     Joel
    -0.43
     Jaffe
    -0.43
     «
    -0.42
     Jacobson
    -0.42
    POSITIVE LOGITS
     Think
    1.33
    Think
    1.32
     THINK
    1.30
     think
    1.27
    think
    1.25
    THINK
    1.21
     thinks
    1.16
     Thinking
    1.04
     thinking
    1.02
     thought
    0.98
    Act Density 0.080%

    No Known Activations