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

    Neuronpedia

    INDEX
    Explanations

    Wikipedia articles

    np_max-act · gemini-2.0-flash

    The neuron is effectively dead—it never activates on any token and thus isn’t detecting any feature.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     introduce
    -0.08
     transcripts
    -0.07
     ráno
    -0.07
     SPL
    -0.07
    International
    -0.06
     International
    -0.06
     svým
    -0.06
     international
    -0.06
     metav
    -0.06
    /Create
    -0.06
    POSITIVE LOGITS
     sqlCommand
    0.07
     ONLY
    0.07
     rol
    0.07
            ↵↵
    0.07
    ãeste
    0.07
    annon
    0.06
    833
    0.06
    _brand
    0.06
                    ↵↵
    0.06
     애
    0.06
    Act Density 0.004%

    No Known Activations