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

    Neuronpedia

    INDEX
    Explanations

    benchmarks and experimental results

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     appren
    0.44
    信息的
    0.40
    aprend
    0.38
    เรียน
    0.38
    ीत
    0.37
     begrij
    0.37
    엘
    0.37
     renounce
    0.36
     एडज
    0.36
    ච
    0.36
    POSITIVE LOGITS
     benchmarks
    1.16
     benchmark
    1.13
     benchmarking
    1.03
     evaluation
    1.01
    benchmark
    0.99
     evaluations
    0.97
     experiments
    0.95
     comparison
    0.94
     comparisons
    0.94
     evaluated
    0.92
    Act Density 0.033%

    No Known Activations