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

    Neuronpedia

    INDEX
    Explanations

    special characters and syntax commonly used in programming or code

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

    code assignments and definitions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Staates
    -0.25
     and
    -0.22
     conoció
    -0.21
     during
    -0.20
     who
    -0.20
     were
    -0.20
    ,
    -0.19
     вот
    -0.19
     whose
    -0.19
     also
    -0.19
    POSITIVE LOGITS
     kasarigan
    1.02
    ſicht
    0.94
     témoig
    0.92
    <unused41>
    0.92
    <unused28>
    0.91
    <unused8>
    0.91
    [@BOS@]
    0.91
    <unused14>
    0.91
    <unused3>
    0.91
    <pad>
    0.91
    Act Density 0.011%

    No Known Activations