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

    Neuronpedia

    INDEX
    Explanations

    sentences where a person is being described or mentioned along with a trait or achievement

    oai_token-act-pair · gpt-3.5-turbo

    references to achievements or accomplishments

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    Weakly activates in any phrase containing “a”

    Explanation Uploaded by User
    New Auto-Interp
    Negative Logits
    evidence
    -0.86
    grounds
    -0.76
    Events
    -0.76
    Rules
    -0.74
    inventoryQuantity
    -0.73
    lords
    -0.72
     Actions
    -0.71
    peak
    -0.71
    casts
    -0.69
    events
    -0.68
    POSITIVE LOGITS
     lot
    1.04
     handful
    1.04
     stint
    1.01
     lifelong
    1.00
     whopping
    1.00
     girlfriend
    0.99
     lengthy
    0.99
     suitcase
    0.94
     dozen
    0.94
     brief
    0.94
    Act Density 0.385%

    No Known Activations