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

    Neuronpedia

    INDEX
    Explanations

    key decision-making phrases and expressions related to choices and preferences

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

    prefer / prefolds

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     ſch
    -0.55
     quæ
    -0.52
     ſtate
    -0.46
     ſche
    -0.46
     houſe
    -0.41
     itſelf
    -0.38
     ſtre
    -0.38
    DataSetChanged
    -0.38
     purpoſe
    -0.38
     monaster
    -0.37
    POSITIVE LOGITS
    bootstrapcdn
    0.64
    AndEndTag
    0.63
     preferring
    0.62
    msgTypes
    0.59
     mainly
    0.58
     primarily
    0.57
     particularly
    0.57
     preferable
    0.55
     prefer
    0.55
     prefers
    0.54
    Act Density 0.040%

    No Known Activations