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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating potential for improvement and lessons learned

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

    "improvement" or things to improve upon

    np_acts-logits-general · gemini-2.0-flash

    improvement

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Ecotoxicity
    -0.45
    onavir
    -0.41
    astéroïdes
    -0.40
    omock
    -0.36
    ❹
    -0.36
    ネタバレ
    -0.35
     utafitiHapana
    -0.35
    ◝
    -0.35
    ruma
    -0.35
     railroads
    -0.35
    POSITIVE LOGITS
     improvement
    0.60
    improvement
    0.59
     improvements
    0.57
    /**
    0.56
     mejoras
    0.56
    RetentionPolicy
    0.55
     melh
    0.54
    amélioration
    0.54
    Improvements
    0.52
     Improvement
    0.51
    Act Density 0.378%

    No Known Activations