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

    Neuronpedia

    INDEX
    Explanations

    conjunctive phrases indicating conclusions or logical connections in text

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

    discourse connectives and conjunctions that link clauses or ideas together.

    oai_token-act-pair · claude-4-5-haikuTriggered by @emiglarou
    New Auto-Interp
    Negative Logits
     however
    -0.75
     because
    -0.75
    because
    -0.73
     bowiem
    -0.70
    因为
    -0.68
     ведь
    -0.67
    however
    -0.63
     porque
    -0.63
     totiž
    -0.63
     però
    -0.61
    POSITIVE LOGITS
    จึง
    0.78
     naturally
    0.75
    只好
    0.73
    必然
    0.70
     correspondingly
    0.68
     accordingly
    0.68
     consequently
    0.66
    余計
    0.66
     logically
    0.66
     necessarily
    0.63
    Act Density 1.009%

    No Known Activations