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

    Neuronpedia

    INDEX
    Explanations

    Code snippets

    np_max-act · gemini-2.0-flash

    This neuron detects the structural JSON/Jupyter‐notebook markup (e.g. quotation marks, braces, and key names like “cells” or “source”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    rition
    -0.07
     prediction
    -0.07
     Fabric
    -0.06
     currently
    -0.06
    cx
    -0.06
    _closed
    -0.06
    RequestMapping
    -0.06
     Shelter
    -0.06
    :'
    -0.06
    weet
    -0.06
    POSITIVE LOGITS
     pcm
    0.07
     preocup
    0.06
     раді
    0.06
    %/
    0.06
     автомати
    0.06
    \Auth
    0.06
    .about
    0.06
     heiß
    0.06
    _tp
    0.06
    nullptr
    0.06
    Act Density 0.010%

    No Known Activations