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

    Neuronpedia

    INDEX
    Explanations

    observation

    np_max-act · gemini-2.0-flash

    This neuron never activates—it effectively looks for nothing (it’s a “dead” neuron).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    .Formatting
    -0.06
     depot
    -0.06
    -mean
    -0.06
    RESET
    -0.06
     dust
    -0.06
     IPs
    -0.06
    ,tp
    -0.06
    _ca
    -0.06
    Duplicates
    -0.06
     dusty
    -0.06
    POSITIVE LOGITS
     observation
    0.11
    observation
    0.10
     observations
    0.08
    observations
    0.08
     Observation
    0.08
     incur
    0.07
     athleticism
    0.07
    巴
    0.06
     observe
    0.06
    )(↵
    0.06
    Act Density 0.017%

    No Known Activations