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

    Neuronpedia

    INDEX
    Explanations

    Suffixes

    np_max-act · gemini-2.0-flash

    The neuron activates on hyphen-prefixed single-character segments, i.e. hyphenated affix markers in compound words.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Compar
    -0.07
     expelled
    -0.07
     Kamp
    -0.07
     baked
    -0.07
     consul
    -0.06
    ومان
    -0.06
     Braun
    -0.06
     часов
    -0.06
    SPELL
    -0.06
     ingin
    -0.06
    POSITIVE LOGITS
    	↵	↵↵
    0.07
    -z
    0.06
    -Y
    0.06
    -r
    0.06
    agua
    0.06
     такая
    0.06
     NSError
    0.06
     Alright
    0.06
      ↵↵↵
    0.06
    <*
    0.06
    Act Density 0.043%

    No Known Activations