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

    Neuronpedia

    INDEX
    Explanations

    Highs and lows

    np_max-act · gemini-2.0-flash

    conversation-related text, particularly in interactive dialogues or customer service contexts.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    This neuron detects comparative and superlative degree words (e.g. “most,” “more,” “highest,” “advanced”) indicating relative scale or emphasis.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Yugoslavia
    -0.06
     fidelity
    -0.06
     gioc
    -0.06
     attends
    -0.06
     OCT
    -0.06
    xxxxxxxx
    -0.06
    _subscription
    -0.06
     сиг
    -0.06
     think
    -0.06
    =-=-=-=-
    -0.06
    POSITIVE LOGITS
     마음
    0.08
    //--------------------------------------------------------------↵
    0.07
     NDEBUG
    0.07
     AVAILABLE
    0.07
    callable
    0.07
    quarter
    0.06
     MethodInvocation
    0.06
     extraordin
    0.06
     hyperlink
    0.06
    leine
    0.06
    Act Density 0.039%

    No Known Activations