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

    Neuronpedia

    INDEX
    Explanations

    multiple languages

    np_max-act · gemini-2.0-flash

    The neuron is primarily detecting subword tokens containing French diacritical marks (accented characters).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     dosp
    -0.07
     SHOW
    -0.06
     Pam
    -0.06
     Kush
    -0.06
     Status
    -0.06
     spiders
    -0.06
    ंतर
    -0.06
     foul
    -0.06
    unkt
    -0.06
    Interesting
    -0.06
    POSITIVE LOGITS
    mainwindow
    0.07
    Supported
    0.07
     rua
    0.06
    Textures
    0.06
    ész
    0.06
    .nz
    0.06
    	Resource
    0.06
    .empty
    0.06
     axs
    0.06
     deceased
    0.06
    Act Density 0.006%

    No Known Activations