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

    Neuronpedia

    INDEX
    Explanations

    choosing

    np_max-act · gemini-2.0-flash

    The neuron detects multiple-choice question prompts—specifically phrases instructing you to “choose one [letter/option] from” the given options.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    conditional reasoning or logical conclusions based on premises presented.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    (completion
    -0.06
     LAP
    -0.06
    、な
    -0.06
     Performs
    -0.06
    _OK
    -0.06
    .servers
    -0.06
    植
    -0.06
     mesaj
    -0.06
    若
    -0.06
    .mag
    -0.06
    POSITIVE LOGITS
    imeo
    0.07
     getUrl
    0.07
     FIFA
    0.06
    .locale
    0.06
     investing
    0.06
    paramref
    0.06
    阶段
    0.06
    _EVAL
    0.06
     TokenType
    0.06
     Partnership
    0.06
    Act Density 0.009%

    No Known Activations