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

    Neuronpedia

    INDEX
    Explanations

    phrases related to pulling an object or executing a successful action

    oai_token-act-pair · gpt-3.5-turbo

    phrases involving the actions of pulling or lifting

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    externalActionCode
    -0.84
    anwhile
    -0.81
    nown
    -0.75
    ailability
    -0.73
    haven
    -0.69
    umbered
    -0.67
    FACE
    -0.65
    ufact
    -0.64
    da
    -0.63
    lia
    -0.62
    POSITIVE LOGITS
     punches
    0.89
     levers
    0.82
     cart
    0.82
     carts
    0.81
     scrimmage
    0.76
     lever
    0.73
    ©¶æ
    0.72
     straw
    0.69
     wedge
    0.67
     wagon
    0.66
    Act Density 0.164%

    No Known Activations