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

    Neuronpedia

    INDEX
    Explanations

    instances where someone is declining to comment

    oai_token-act-pair · gpt-3.5-turbo

    instances of refusal to comment

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    Offline
    -0.74
    ICAN
    -0.72
    tek
    -0.72
    aptic
    -0.70
    sworth
    -0.70
    viks
    -0.68
    Haunted
    -0.68
    horse
    -0.68
    neys
    -0.67
    bags
    -0.67
    POSITIVE LOGITS
     Krug
    0.70
    otiation
    0.69
     predec
    0.68
     admission
    0.68
     invitations
    0.68
     commission
    0.67
     reimbursement
    0.66
     sharply
    0.66
     unanimously
    0.66
     confir
    0.65
    Act Density 0.019%

    No Known Activations