README.md

---
dataset_info:
  features:
  - name: experiment_id
    dtype: int64
  - name: step
    dtype: int64
  - name: sequence
    dtype: string
  - name: activation
    dtype: float64
  - name: metric_value
    dtype: float64
  - name: label
    dtype: string
  - name: pos_match_string
    dtype: string
  - name: neg_match_string
    dtype: string
  splits:
  - name: train
    num_bytes: 380146
    num_examples: 1408
  download_size: 165074
  dataset_size: 380146
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---