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

    Neuronpedia

    INDEX
    Explanations

    instances of the word "when" indicating temporal references

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    agin
    -0.73
    Die
    -0.70
    omal
    -0.70
    urden
    -0.69
    icker
    -0.68
    lation
    -0.66
    andal
    -0.66
    ictive
    -0.65
    odge
    -0.63
    abre
    -0.63
    POSITIVE LOGITS
     discussing
    0.89
     asked
    0.83
     he
    0.81
     faced
    0.78
     confronted
    0.77
     researching
    0.77
     introducing
    0.75
     they
    0.74
     encountering
    0.74
     pressed
    0.72
    Act Density 0.065%

    No Known Activations