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

    Neuronpedia

    INDEX
    Explanations

    rate

    np_max-act · gemini-2.0-flash

    techniques for evaluating the performance of large language models.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     pack
    -0.07
    onyms
    -0.06
    	login
    -0.06
     말
    -0.06
    olleyError
    -0.06
    $email
    -0.06
     Lansing
    -0.06
     Republicans
    -0.06
     CFG
    -0.06
    fs
    -0.06
    POSITIVE LOGITS
     harsh
    0.07
     оцен
    0.07
    ody
    0.07
    _od
    0.07
    _HI
    0.07
    序
    0.06
    (Method
    0.06
    >B
    0.06
    лож
    0.06
     INS
    0.06
    Act Density 0.015%

    No Known Activations