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

    Neuronpedia

    INDEX
    Explanations

    programmed to avoid

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     carico
    0.38
    দীর্ঘ
    0.38
    焼
    0.38
    /**
    0.37
     outsourced
    0.37
     conseg
    0.36
    Legacy
    0.36
     uniquely
    0.36
    获得了
    0.36
     tasar
    0.36
    POSITIVE LOGITS
     avoid
    0.90
     evitar
    0.90
    避免
    0.84
     evitare
    0.84
     menghindari
    0.81
     éviter
    0.79
     evitando
    0.78
    avoid
    0.77
     tránh
    0.77
     refrained
    0.77
    Act Density 0.318%

    No Known Activations