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

    Neuronpedia

    INDEX
    Explanations

    Formal/technical writing

    np_max-act · gemini-2.0-flash

    synthetic chemical compounds and their production methods in the chemical industry.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    The neuron fires on the words and punctuation that introduce or number individual synthesis routes (e.g. “One,” list‐item markers, and the dashes in “2‐methyl” etc.), i.e. the tokens that label and start each step or route description.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ER
    -0.07
     mnohem
    -0.06
     }];↵↵
    -0.06
     mover
    -0.06
    オ
    -0.06
     SPA
    -0.06
    PLIED
    -0.06
     DOCUMENT
    -0.06
    WER
    -0.06
    (REG
    -0.06
    POSITIVE LOGITS
    229
    0.07
    apid
    0.06
    ///
↵
    0.06
    něž
    0.06
    Others
    0.06
    .bit
    0.06
    should
    0.06
    usses
    0.06
    lust
    0.06
     свидетель
    0.06
    Act Density 0.054%

    No Known Activations