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

    Neuronpedia

    INDEX
    Explanations

    words related to connecting or merging

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "connect" and related forms indicating relationships or links between entities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Leaves
    -0.71
    cheat
    -0.66
    hra
    -0.66
    Ħ¢
    -0.64
    corn
    -0.64
    birth
    -0.63
     spoil
    -0.63
     Ambrose
    -0.60
    basketball
    -0.59
     restraining
    -0.58
    POSITIVE LOGITS
    icut
    1.01
     dots
    1.00
     Connect
    0.89
     connections
    0.85
    nect
    0.85
    links
    0.83
    Connect
    0.83
    anooga
    0.82
    ioned
    0.81
    link
    0.80
    Act Density 0.022%

    No Known Activations