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

    Neuronpedia

    INDEX
    Explanations

    70-90

    np_max-act · gemini-2.0-flash

    The neuron primarily detects numeric decade references (e.g. “60,” “80,” “90”) and their “s” suffixes indicating the 60s, 80s, 90s, etc.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    REQUIRED
    -0.07
     Resource
    -0.07
    =df
    -0.07
    ncmp
    -0.07
    .ToUpper
    -0.07
    ())/
    -0.07
     thus
    -0.07
    strncmp
    -0.07
     systemd
    -0.07
     TripAdvisor
    -0.07
    POSITIVE LOGITS
     moms
    0.07
     publishing
    0.07
     thirty
    0.07
     hairy
    0.06
     seventy
    0.06
     mom
    0.06
     Shield
    0.06
     fifty
    0.06
     Thirty
    0.06
     meals
    0.06
    Act Density 0.005%

    No Known Activations