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

    Neuronpedia

    INDEX
    Explanations

    phrases related to exertion or difficulty

    oai_token-act-pair · gpt-3.5-turbo

    phrases that indicate being challenged or under pressure

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     succession
    -0.70
     dynamics
    -0.65
     legitimacy
    -0.60
     occurrence
    -0.59
     originate
    -0.59
     characteristic
    -0.58
     CTR
    -0.58
     divergence
    -0.57
    VERTISEMENT
    -0.57
     spurious
    -0.57
    POSITIVE LOGITS
    pressed
    0.90
    angering
    0.79
    icent
    0.78
    cerned
    0.78
    ped
    0.75
    eled
    0.75
    ivated
    0.75
    watching
    0.74
    nin
    0.73
    oing
    0.73
    Act Density 0.483%

    No Known Activations