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

    Neuronpedia

    INDEX
    Explanations

    Games

    np_max-act · gemini-2.0-flash

    This neuron detects mentions of the game name “Battleship,” including its subword fragments (e.g. “battles” + “hip”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    /feed
    -0.06
     �
    -0.06
    wife
    -0.06
     Rams
    -0.06
    904
    -0.06
     attends
    -0.06
     �
    -0.06
     TableView
    -0.06
     dungeon
    -0.06
    .Pe
    -0.06
    POSITIVE LOGITS
    ินการ
    0.07
    .invoke
    0.07
     communicating
    0.07
     forControlEvents
    0.07
    での
    0.06
     پای
    0.06
     vieux
    0.06
     continental
    0.06
     �
    0.06
    ahir
    0.06
    Act Density 0.033%

    No Known Activations