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

    Neuronpedia

    INDEX
    Explanations

    emotional adjectives that express positive or negative feelings

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

    words ending in -ful

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     incest
    -0.57
     avoient
    -0.57
     mijne
    -0.56
     treason
    -0.54
     hjemmeside
    -0.53
     homosexuality
    -0.52
     amnesty
    -0.52
     bribe
    -0.52
     mosque
    -0.52
     kolye
    -0.52
    POSITIVE LOGITS
    eful
    0.68
    Peaceful
    0.66
    neurial
    0.65
     Peaceful
    0.64
    Stateful
    0.63
    graceful
    0.63
     mindful
    0.63
     Faithful
    0.62
    peaceful
    0.62
    ropical
    0.61
    Act Density 0.017%

    No Known Activations