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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating potential solutions and improvements to problems

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

    followed by "this" or similar terms

    np_acts-logits-general · gemini-2.0-flash

    achieving goals, solving problems, providing answers

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    parsedMessage
    -0.84
    ьаж
    -0.82
    Бахар
    -0.81
    [@BOS@]
    -0.80
    <unused43>
    -0.79
    <unused42>
    -0.79
    <unused51>
    -0.79
    <unused52>
    -0.79
    <unused3>
    -0.79
    <unused8>
    -0.79
    POSITIVE LOGITS
     these
    0.53
     this
    0.49
     tersebut
    0.43
     such
    0.42
    そんな
    0.41
     diesen
    0.40
    这种情况
    0.38
    这些
    0.38
     diese
    0.37
     چنین
    0.37
    Act Density 0.642%

    No Known Activations