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

    Neuronpedia

    INDEX
    Explanations

    identifying word completions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     detract
    0.41
    ientras
    0.38
    requently
    0.37
     overheard
    0.36
    <unused2164>
    0.36
    $:
    0.36
     किसी
    0.35
     जबकि
    0.35
     diminishes
    0.35
    .|__
    0.35
    POSITIVE LOGITS
    🥹
    0.43
    様専用
    0.41
     ransomware
    0.41
     Sehen
    0.36
    🫧
    0.36
     eSIM
    0.35
    🫶
    0.35
    -
    0.35
    0.34
     bitcoin
    0.34
    Act Density 0.252%

    No Known Activations