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

    Neuronpedia

    INDEX
    Explanations

    words related to having an advantage or being at the forefront

    oai_token-act-pair · gpt-3.5-turbo

    references to the concept of "edge" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ania
    -0.77
    tics
    -0.74
    asca
    -0.73
    adult
    -0.72
    HCR
    -0.70
    anism
    -0.69
    ractor
    -0.69
    oney
    -0.67
    sen
    -0.67
    oric
    -0.67
    POSITIVE LOGITS
     edge
    1.12
     Edge
    0.94
     edges
    0.88
     Catalyst
    0.87
    Edge
    0.84
     rusher
    0.81
     fence
    0.74
     margin
    0.74
    hog
    0.74
     breaker
    0.65
    Act Density 0.009%

    No Known Activations