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

    Neuronpedia

    INDEX
    Explanations

    words related to secrets or confidential matters

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "confide" and its variations, indicating themes of trust or confidentiality

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    senal
    -0.85
    hyde
    -0.84
    DAY
    -0.80
    BALL
    -0.76
    manship
    -0.72
    ï¸ı
    -0.71
    schild
    -0.69
    ForgeModLoader
    -0.67
    NING
    -0.67
    ©¶æ
    -0.66
    POSITIVE LOGITS
    luence
    1.17
    essional
    1.13
    ederation
    1.11
    eder
    1.05
    usions
    1.01
    licted
    0.94
    eree
    0.92
    etti
    0.89
    usional
    0.87
    essor
    0.86
    Act Density 0.005%

    No Known Activations