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

    Neuronpedia

    INDEX
    Explanations

    pursuing human-defined goals

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Размер
    0.93
     Featuring
    0.91
     сегодняшний
    0.90
     Wonders
    0.88
     Examine
    0.87
     collectionView
    0.87
    Rustic
    0.87
     Texte
    0.87
    تاریخ
    0.86
     culprits
    0.85
    POSITIVE LOGITS
     subgoal
    1.23
     optimality
    1.10
     heuristics
    1.04
     optimally
    1.04
     useful
    1.01
     Bayesian
    0.98
     suboptimal
    0.96
     autonomously
    0.95
     optimization
    0.95
     rationally
    0.93
    Act Density 0.190%

    No Known Activations