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

    Neuronpedia

    INDEX
    Explanations

    warnings, mistakes, unfortunate events

    np_max-act · gemini-2.0-flash

    This neuron fires on regretful or reproachful “should have known better (than to …)” constructions that express someone ought to have anticipated or avoided something.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     boring
    -0.07
    dfs
    -0.07
    ِين
    -0.07
     decis
    -0.07
    _ENTER
    -0.06
    xiety
    -0.06
    _word
    -0.06
    838
    -0.06
    iris
    -0.06
     papers
    -0.06
    POSITIVE LOGITS
    -Control
    0.07
    <UnityEngine
    0.06
     Guaranteed
    0.06
     Indicates
    0.06
     perv
    0.06
     Cumhur
    0.06
    (Function
    0.06
    _Box
    0.06
    (Application
    0.06
     demonstrated
    0.06
    Act Density 0.030%

    No Known Activations