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

    Neuronpedia

    INDEX
    Explanations

    London

    np_max-act · gemini-2.0-flash

    The neuron flags tokens that occur immediately before numeric quantities (i.e. it activates on words directly preceding numbers or measurements).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Throwable
    -0.07
     Men
    -0.06
     Orn
    -0.06
     Padres
    -0.06
     Terr
    -0.06
     Fonts
    -0.06
     Sid
    -0.06
     Ru
    -0.06
     Bedrooms
    -0.06
     nr
    -0.06
    POSITIVE LOGITS
     architecture
    0.08
     rhythm
    0.07
    github
    0.07
    _identity
    0.06
    _cursor
    0.06
     electric
    0.06
    �인
    0.06
    ことに
    0.06
    调用
    0.06
    ocom
    0.06
    Act Density 0.346%

    No Known Activations