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

    Neuronpedia

    INDEX
    Explanations

    words related to physical actions or movements, particularly those involving physical contact or struggle

    oai_token-act-pair · gpt-3.5-turbo

    instances of words ending in 'les', 'led', or 'ling', indicating a focus on certain word forms

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ĻĤ
    -0.86
    ÑĢ
    -0.73
    ãĥ¼ãĥ³
    -0.68
    EMP
    -0.67
    ITNESS
    -0.65
    л
    -0.65
    urd
    -0.64
    Inf
    -0.63
    OGR
    -0.63
    raviolet
    -0.61
    POSITIVE LOGITS
    puff
    1.12
    ptin
    0.98
    cone
    0.90
    tail
    0.85
    jee
    0.79
    iness
    0.78
    kered
    0.77
     awkwardly
    0.76
    ishly
    0.76
     nervously
    0.74
    Act Density 0.100%

    No Known Activations