tim-lawson/sae-pythia-160m-deduped-x64-k32-tfm-layers-10-dists

该数据集包含多个特征字段，包括latent（潜在变量）、count（计数）、total（总计）、mean（平均值）、layer_mean（层级平均值）、layer_var（层级方差）、layer_std（层级标准差）、counts（计数序列）、totals（总计序列）、means（平均值序列）和probs（概率序列）。数据集被分割为训练集，包含49152个样本，总大小为22413312字节。

This dataset includes multiple feature fields such as latent (latent variable), count (count), total (total), mean (average), layer_mean (layer average), layer_var (layer variance), layer_std (layer standard deviation), counts (sequence of counts), totals (sequence of totals), means (sequence of averages), and probs (sequence of probabilities). The dataset is split into a training set containing 49152 samples, with a total size of 22413312 bytes.