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

    Neuronpedia

    INDEX
    Explanations

    attends to the latest closing bracket from arbitrary tokens related to citations or references

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.06
    1:0.08
    2:0.05
    3:0.07
    4:0.11
    5:0.06
    6:0.38
    7:0.14
    Negative Logits
    AndEndTag
    -0.51
    OGND
    -0.50
     Paglinawan
    -0.49
    tonsoft
    -0.45
     bezeichneter
    -0.45
     виправивши
    -0.45
    parsedMessage
    -0.45
     betweenstory
    -0.44
     استنادى
    -0.43
     MessageBoxIcon
    -0.43
    POSITIVE LOGITS
     N
    0.29
     He
    0.25
    =
    0.23
     W
    0.22
     U
    0.22
     R
    0.22
     reference
    0.22
    N
    0.21
     A
    0.21
     M
    0.21
    Act Density 1.050%

    No Known Activations