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

    Neuronpedia

    INDEX
    Explanations

    media and entertainment

    np_max-act · gemini-2.0-flash

    The neuron fires most strongly on standalone “topic” nouns—single words naming the main object or medium in a passage (e.g. series, movie, show, game, spreadsheet, assembly, boat).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    >());↵
    -0.07
    .IntegerField
    -0.07
     wel
    -0.07
     diferentes
    -0.06
    }`,↵
    -0.06
     egregious
    -0.06
    	namespace
    -0.06
    ',['../
    -0.06
    ];↵
    -0.06
    but
    -0.06
    POSITIVE LOGITS
    _logs
    0.09
     Phil
    0.07
     LSD
    0.07
     shredded
    0.06
     dislikes
    0.06
     cops
    0.06
    olders
    0.06
     DSP
    0.06
    雨
    0.06
     noticed
    0.06
    Act Density 0.052%

    No Known Activations