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

    Neuronpedia

    INDEX
    Explanations

    obstacles and hindrance

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on sub-word pieces of gerunds and present‐participle verbs (i.e. “-ing” forms).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    _identity
    -0.07
    wrong
    -0.06
    GetProperty
    -0.06
     Maths
    -0.06
     copyrighted
    -0.06
    _credentials
    -0.06
    _pcm
    -0.06
     conserve
    -0.05
    _pf
    -0.05
    _resp
    -0.05
    POSITIVE LOGITS
     barriers
    0.10
     imped
    0.09
     obstacles
    0.09
     hinder
    0.08
     obstacle
    0.08
    details
    0.08
     prohibiting
    0.07
     Hind
    0.07
     suppress
    0.07
     muddy
    0.07
    Act Density 0.024%

    No Known Activations