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

    Neuronpedia

    INDEX
    Explanations

    problem/solution

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     domain
    0.45
     exchanges
    0.40
    domain
    0.39
    域
    0.38
     invites
    0.37
     tweak
    0.36
     accepts
    0.36
     envis
    0.36
     mappings
    0.35
     transcripts
    0.35
    POSITIVE LOGITS
     Sleep
    0.44
     Podczas
    0.42
    idle
    0.41
    বর্তমানে
    0.41
     присут
    0.41
     żad
    0.41
    тири
    0.41
    nikiem
    0.41
     点
    0.40
     poważ
    0.40
    Act Density 0.002%

    No Known Activations