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

    Neuronpedia

    INDEX
    Explanations

    phrases related to the ease or difficulty of tasks and processes

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

    difficulty and ease

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Мексичка
    -0.81
     beſch
    -0.78
    المناصب
    -0.78
     deſſen
    -0.75
     surla
    -0.75
     geſch
    -0.74
     nakalista
    -0.73
    ➟
    -0.72
     ſeines
    -0.72
    [@BOS@]
    -0.71
    POSITIVE LOGITS
     due
    0.37
     because
    0.36
     greatly
    0.34
     He
    0.33
     easy
    0.32
     great
    0.32
     and
    0.31
     everywhere
    0.31
    easy
    0.31
    .
    0.30
    Act Density 0.054%

    No Known Activations