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

    Neuronpedia

    INDEX
    Explanations

    concessive clauses

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     zodat
    0.39
     takže
    0.36
     sodass
    0.33
     gdyż
    0.31
     ताकि
    0.30
     waardoor
    0.28
     deoarece
    0.27
     જેથી
    0.26
    但这
    0.26
     ይህም
    0.26
    POSITIVE LOGITS
     however
    0.40
     meanwhile
    0.36
    however
    0.33
     unsurprisingly
    0.32
     unlike
    0.30
     viewed
    0.30
     driven
    0.29
     upon
    0.28
     supported
    0.28
     in
    0.27
    Act Density 0.062%

    No Known Activations