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

    Neuronpedia

    INDEX
    Explanations

    occurrences of the verb "to do" in various forms

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

    "does" or "did" followed by "not"

    np_acts-logits-general · gemini-2.0-flash

    does not or did not

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     <<<<<<<<<<<<<<
    -0.66
     disambiguazione
    -0.58
    RegressionTest
    -0.51
     LRT
    -0.50
    StackNavigator
    -0.49
    intellij
    -0.46
     forKey
    -0.45
     Branca
    -0.45
     pref
    -0.44
    centerTitle
    -0.43
    POSITIVE LOGITS
     does
    1.15
     did
    1.05
    does
    1.01
     Does
    0.96
    did
    0.93
     do
    0.92
    Does
    0.89
     Did
    0.85
    Did
    0.85
     DOES
    0.81
    Act Density 0.144%

    No Known Activations