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

    Neuronpedia

    INDEX
    Explanations

    TI6

    np_acts-logits-general · gemini-2.5-flash-lite

    This neuron detects numeric tokens—numbers, years, scores, and other digit-containing tokens in the text.

    oai_token-act-pair · gpt-5-miniTriggered by @yooniel31
    New Auto-Interp
    Negative Logits
     formalism
    0.34
     hysteresis
    0.34
     deleterious
    0.33
    herence
    0.32
     passivation
    0.32
     paraphr
    0.32
     interfacial
    0.31
     astring
    0.31
    zelfde
    0.31
     multivalued
    0.31
    POSITIVE LOGITS
     Kochi
    0.30
     約
    0.29
     April
    0.29
     DiCaprio
    0.29
     올해
    0.29
    September
    0.28
     레
    0.28
     yaklaşık
    0.28
    、
    0.28
    🥇
    0.28
    Act Density 0.069%

    No Known Activations