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

    Neuronpedia

    INDEX
    Explanations

    code

    np_max-act · gemini-2.0-flash

    queries related to modifying or retrieving database records.

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

    The neuron detects words and phrases that signal problems or failures (e.g. “but,” “don’t,” “working”)—i.e. expressions of code not working.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Beautiful
    -0.06
    disabled
    -0.06
     addictive
    -0.06
    erves
    -0.06
    	selected
    -0.06
     testimony
    -0.05
     Slow
    -0.05
     cured
    -0.05
    _Current
    -0.05
     İyi
    -0.05
    POSITIVE LOGITS
     timp
    0.07
     GNU
    0.07
    acje
    0.07
     crushers
    0.07
    ności
    0.07
    (numero
    0.06
    .DataSource
    0.06
     isFirst
    0.06
     پیوند
    0.06
    losti
    0.06
    Act Density 0.051%

    No Known Activations