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

    Neuronpedia

    INDEX
    Explanations

    isolation

    np_max-act · gemini-2.0-flash

    The neuron fires strongly on tokens that indicate being alone or left by oneself (e.g. “sozinhos”→so z inh os in Portuguese, “alone,” “with,” “in” in context of isolation). In other words, it detects mentions of a subject being alone or isolated in a location.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Lydia
    -0.07
     أ
    -0.07
     Jade
    -0.06
    Victoria
    -0.06
     애
    -0.06
     Renaissance
    -0.06
    .include
    -0.06
     Martins
    -0.06
    เลย
    -0.06
    .CL
    -0.06
    POSITIVE LOGITS
    controlled
    0.07
     вид
    0.07
     internet
    0.06
     perceived
    0.06
     Ran
    0.06
     fantasies
    0.06
    _host
    0.06
    activ
    0.06
     *@
    0.06
    abis
    0.06
    Act Density 0.008%

    No Known Activations