INDEX
    Explanations

    metal-related words and items

    New Auto-Interp
    Negative Logits
     [|
    -0.87
    zee
    -0.72
    ortion
    -0.70
    inen
    -0.70
    NEY
    -0.65
    SPONSORED
    -0.64
    orsi
    -0.64
    YA
    -0.64
    oji
    -0.63
     Kard
    -0.63
    POSITIVE LOGITS
    anguage
    1.20
    clad
    1.13
     oxide
    1.10
    works
    1.07
    fish
    1.01
     alloy
    1.00
     shards
    0.99
     ore
    0.96
    heads
    0.96
     flakes
    0.95
    Act Density 2.441%

    No Known Activations