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

    Neuronpedia

    INDEX
    Explanations

    Hacker News IDs

    np_max-act · gemini-2.0-flash

    references to technology-related discussions and products.

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

    The neuron selectively activates on standalone numeric tokens (integers or decimals), especially those appearing as URL query parameters or identifiers.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Wr
    -0.07
     Russo
    -0.07
     Л
    -0.06
     Ρ
    -0.06
     mun
    -0.06
     hab
    -0.06
    .Controls
    -0.06
    άνα
    -0.06
    Pages
    -0.06
     atroc
    -0.06
    POSITIVE LOGITS
     commit
    0.07
     refugee
    0.06
    je
    0.06
    这种
    0.06
    (elem
    0.06
    Market
    0.06
    ريف
    0.06
     NSMutableDictionary
    0.06
     motivational
    0.06
    Messenger
    0.06
    Act Density 0.001%

    No Known Activations