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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating effectiveness or the outcome of actions

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

    how something works in practice

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     avoient
    -0.56
     monasterio
    -0.48
    ⓧ
    -0.48
     CreateTagHelper
    -0.47
     étoit
    -0.46
    etermined
    -0.44
     juſ
    -0.43
    +-+-
    -0.41
    principalColumn
    -0.39
     itſelf
    -0.39
    POSITIVE LOGITS
     effectively
    1.12
    effectively
    1.11
     effective
    1.05
     Effectively
    1.05
    Effective
    1.03
    effective
    1.02
     Effective
    1.00
    Effec
    0.94
     EFFECTIVE
    0.88
     efektif
    0.81
    Act Density 0.171%

    No Known Activations