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

    Neuronpedia

    INDEX
    Explanations

    margins and blocking

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     hallucinations
    0.42
    [{\
    0.42
     aromas
    0.40
     kaleidoscopic
    0.40
     abstracts
    0.40
    ஷ்ட
    0.40
     abstrak
    0.40
     photons
    0.39
     abstract
    0.38
     aphids
    0.38
    POSITIVE LOGITS
    MARGIN
    0.42
    Margin
    0.42
    blocked
    0.40
    Blocking
    0.40
     reassuring
    0.39
     Margin
    0.39
    refresh
    0.39
    kej
    0.38
    margins
    0.38
    Blocked
    0.37
    Act Density 0.000%

    No Known Activations