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

    Neuronpedia

    INDEX
    Explanations

    hypo/hyper conditions

    np_max-act · gemini-2.0-flash

    The neuron activates on words beginning with “hypo-” (e.g. hypoxia, hypotension, hypoglycemia).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     ten
    -0.08
    olders
    -0.08
     renters
    -0.08
     resin
    -0.08
     lance
    -0.08
     replic
    -0.07
     dresser
    -0.07
    OSE
    -0.07
     caliber
    -0.07
     eight
    -0.07
    POSITIVE LOGITS
    hyp
    0.12
     hyp
    0.11
     Hyp
    0.10
     hypocrisy
    0.08
     （
    0.07
     гід
    0.07
    .Top
    0.07
    v
    0.06
     loophole
    0.06
     hypothetical
    0.06
    Act Density 0.011%

    No Known Activations