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

    Neuronpedia

    INDEX
    Explanations

    subtle expressions of opinion or emotional responses

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    Preceding intensifiers such as "really," "very," and "sure."

    np_acts-logits-general · gemini-2.0-flash

    awfully, quite, very, long, sure

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    arXiv
    -0.49
     sinais
    -0.46
     _,
    -0.43
    nader
    -0.42
    зора
    -0.40
    equalsIgnoreCase
    -0.39
    TouchListener
    -0.39
     mema
    -0.39
     aviones
    -0.39
     uniquement
    -0.38
    POSITIVE LOGITS
     awfully
    1.11
     alot
    0.94
     quite
    0.93
     very
    0.91
    quite
    0.90
     pretty
    0.88
    aarrggbb
    0.88
     VERY
    0.87
     certainly
    0.85
     sacré
    0.85
    Act Density 0.273%

    No Known Activations