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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating a direction or inclination

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to directional shifts or leanings

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    nces
    -0.74
    listed
    -0.72
    unte
    -0.70
     Lamb
    -0.65
    enery
    -0.64
    cooked
    -0.64
    ©¶æ¥µ
    -0.64
    bath
    -0.62
    ydia
    -0.62
    NJ
    -0.61
    POSITIVE LOGITS
     toward
    1.02
     favoring
    0.97
     direction
    0.91
     directional
    0.91
     towards
    0.91
     downward
    0.88
     downwards
    0.84
     tilt
    0.84
    wards
    0.81
     focus
    0.78
    Act Density 0.251%

    No Known Activations