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

    Neuronpedia

    INDEX
    Explanations

    comma

    np_max-act · gemini-2.0-flash

    This neuron detects polite invitation or closing phrases (e.g., “please feel free…,” “any concerns…”) that invite further questions.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ret
    -0.06
     Hizmet
    -0.06
     Tank
    -0.06
     applicationWill
    -0.06
    .za
    -0.06
     Estimates
    -0.06
    RET
    -0.06
    Fixture
    -0.06
    preh
    -0.05
    junction
    -0.05
    POSITIVE LOGITS
    Journal
    0.07
    /logs
    0.07
    ımın
    0.07
    _SEQUENCE
    0.06
     chiff
    0.06
    ئيس
    0.06
     але
    0.06
    ひ
    0.06
     الولايات
    0.06
     SendMessage
    0.06
    Act Density 0.017%

    No Known Activations