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

    Neuronpedia

    INDEX
    Explanations

    code

    np_max-act · gemini-2.0-flash

    The neuron fires on the “head” words of two‐word labels or section‐heading compounds (e.g. the second word in “dress code,” “shower house,” “Division of Agriculture,” “Shortcut Action,” etc.).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Verse
    -0.07
    )'),↵
    -0.07
    ��
    -0.06
    	Destroy
    -0.06
    :any
    -0.06
    )this
    -0.06
    icles
    -0.06
    Fluid
    -0.06
    iễ
    -0.06
     Scalars
    -0.06
    POSITIVE LOGITS
    -themed
    0.07
    。また
    0.06
    izzling
    0.06
    emple
    0.06
     estates
    0.06
    Zen
    0.06
    _CYCLE
    0.06
     loved
    0.06
     spread
    0.06
     bottleneck
    0.06
    Act Density 0.000%

    No Known Activations