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

    Neuronpedia

    INDEX
    Explanations

    references to levels of difficulty in games, learning environments, or challenges

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

    task difficulty

    np_acts-logits-general · gemini-2.0-flash

    difficulty levels and challenges

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    onViewCreated
    -0.54
    헌
    -0.53
    kehrt
    -0.46
    ansons
    -0.44
    atzung
    -0.42
    cvt
    -0.41
     mass
    -0.40
    wohn
    -0.40
    ță
    -0.40
     المتحدة
    -0.39
    POSITIVE LOGITS
     difficulty
    2.21
    difficulty
    1.96
     difficult
    1.87
     challenging
    1.86
     Difficulty
    1.85
     Difficult
    1.82
    difficult
    1.79
     Challenging
    1.78
    Difficulty
    1.74
    Difficult
    1.74
    Act Density 0.227%

    No Known Activations