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

    Neuronpedia

    INDEX
    Explanations

    cleaning product reviews

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     separately
    -0.09
     muts
    -0.08
    Vend
    -0.08
     hurting
    -0.08
     ‘‘
    -0.07
     Morris
    -0.07
    Wish
    -0.07
     intervene
    -0.07
    abbo
    -0.07
     wished
    -0.07
    POSITIVE LOGITS
     zuverlässig
    0.14
     zuverläss
    0.12
     efficacy
    0.12
     reliably
    0.10
    效果
    0.10
     gewährleisten
    0.10
     tehok
    0.09
     effectiveness
    0.09
     эффективность
    0.09
    coverage
    0.09
    Act Density 0.135%

    No Known Activations