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

    Neuronpedia

    INDEX
    Explanations

    modal verbs and conditional phrases indicating necessity or possibility

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     St
    -0.55
     kon
    -0.51
     cho
    -0.50
     
    -0.50
     Mon
    -0.48
    坤
    -0.46
     do
    -0.45
     high
    -0.44
     Grand
    -0.43
      
    -0.43
    POSITIVE LOGITS
     meanwhile
    1.25
     itſelf
    1.14
     nevertheless
    1.11
     unfortunately
    1.10
     therefore
    1.07
     neceff
    1.02
     purtroppo
    1.01
     doubtless
    1.00
     certes
    0.99
     however
    0.98
    Act Density 0.667%

    No Known Activations