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

    Neuronpedia

    INDEX
    Explanations

    |

    np_max-act · gemini-2.0-flash

    This neuron detects markdown-style table divider rows (the '---' separator line used to separate headers from table body).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
    -de
    -0.07
    戻
    -0.07
     voleb
    -0.07
    .Mode
    -0.07
    .history
    -0.07
    reffen
    -0.06
    english
    -0.06
     먼
    -0.06
    ační
    -0.06
     APK
    -0.06
    POSITIVE LOGITS
    _symbols
    0.07
     disrupting
    0.06
    quired
    0.06
    [layer
    0.06
    んと
    0.06
     facilitate
    0.06
     "";
↵
    0.06
    促
    0.06
     cycl
    0.06
    _tipo
    0.06
    Act Density 0.004%

    No Known Activations