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

    Neuronpedia

    INDEX
    Explanations

    trend

    np_max-act · gemini-2.0-flash

    This neuron activates on the word “trend” (including compound forms like “trend‐following” or “trend lines”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     ether
    -0.08
     Gabriel
    -0.07
     сахар
    -0.06
    (Uri
    -0.06
     garlic
    -0.06
     Potter
    -0.06
     Mos
    -0.06
    IVER
    -0.06
     boxes
    -0.06
     toile
    -0.06
    POSITIVE LOGITS
     trend
    0.14
     trends
    0.12
     Trends
    0.12
     Trend
    0.11
     trending
    0.09
     Wendy
    0.08
     Brenda
    0.08
     downturn
    0.08
     WN
    0.07
    转
    0.07
    Act Density 0.008%

    No Known Activations