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

    Neuronpedia

    INDEX
    Explanations

    modelmodel checkpointing and training progress

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    resized
    0.46
     dim
    0.45
     dimers
    0.42
     dimensions
    0.41
     Pent
    0.40
    dim
    0.40
    流通
    0.40
    preprocess
    0.40
     dage
    0.38
    bundle
    0.37
    POSITIVE LOGITS
    epoch
    0.76
     epoch
    0.75
    Epoch
    0.73
     Epoch
    0.70
     Monitor
    0.69
    Checkpoint
    0.69
     checkpoints
    0.67
     checkpoint
    0.66
     Monitors
    0.65
    Monitor
    0.64
    Act Density 0.026%

    No Known Activations