Model: GELU Model: 2 Layers, 2048 Neurons per Layer
Dataset: 80% C4 (Web Text) and 20% Python Code
Neuron 14 in Layer 1
Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-2l')
Text #0
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.6603. Min Act: -0.1700
Data Index: 475300 (C4 (Web Text))
Max Activating Token Index: 643
Click toggle to see full text
Truncated
Full Text #0
Text #1
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.5665. Min Act: -0.1700
Data Index: 1209557 (C4 (Web Text))
Max Activating Token Index: 659
Click toggle to see full text
Truncated
Full Text #1
Text #2
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.5353. Min Act: -0.1700
Data Index: 1386506 (Python Code)
Max Activating Token Index: 712
Click toggle to see full text
Truncated
Full Text #2
Text #3
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.4728. Min Act: -0.1700
Data Index: 247606 (C4 (Web Text))
Max Activating Token Index: 698
Click toggle to see full text
Truncated
Full Text #3
Text #4
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.3790. Min Act: -0.1510
Data Index: 1545507 (Python Code)
Max Activating Token Index: 440
Click toggle to see full text
Truncated
Full Text #4
Text #5
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.3790. Min Act: -0.1700
Data Index: 972090 (C4 (Web Text))
Max Activating Token Index: 883
Click toggle to see full text
Truncated
Full Text #5
Text #6
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.3478. Min Act: -0.1700
Data Index: 790656 (C4 (Web Text))
Max Activating Token Index: 77
Click toggle to see full text
Truncated
Full Text #6
Text #7
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.3165. Min Act: -0.1700
Data Index: 855722 (C4 (Web Text))
Max Activating Token Index: 81
Click toggle to see full text
Truncated
Full Text #7
Text #8
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.3165. Min Act: -0.1700
Data Index: 276354 (C4 (Web Text))
Max Activating Token Index: 796
Click toggle to see full text
Truncated
Full Text #8
Text #9
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.2540. Min Act: -0.1700
Data Index: 214400 (C4 (Web Text))
Max Activating Token Index: 123
Click toggle to see full text
Truncated
Full Text #9
Text #10
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.2853. Min Act: -0.1700
Data Index: 1187003 (C4 (Web Text))
Max Activating Token Index: 345
Click toggle to see full text
Truncated
Full Text #10
Text #11
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.2540. Min Act: -0.1700
Data Index: 1599767 (Python Code)
Max Activating Token Index: 985
Click toggle to see full text
Truncated
Full Text #11
Text #12
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.2540. Min Act: -0.1697
Data Index: 1375720 (Python Code)
Max Activating Token Index: 608
Click toggle to see full text
Truncated
Full Text #12
Text #13
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.2853. Min Act: -0.1699
Data Index: 986845 (C4 (Web Text))
Max Activating Token Index: 67
Click toggle to see full text
Truncated
Full Text #13
Text #14
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.2227. Min Act: -0.1700
Data Index: 248861 (C4 (Web Text))
Max Activating Token Index: 250
Click toggle to see full text
Truncated
Full Text #14
Text #15
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.2540. Min Act: -0.1700
Data Index: 285721 (C4 (Web Text))
Max Activating Token Index: 594
Click toggle to see full text
Truncated
Full Text #15
Text #16
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.2227. Min Act: -0.1226
Data Index: 1651049 (Python Code)
Max Activating Token Index: 931
Click toggle to see full text
Truncated
Full Text #16
Text #17
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.2227. Min Act: -0.1700
Data Index: 876119 (C4 (Web Text))
Max Activating Token Index: 564
Click toggle to see full text
Truncated
Full Text #17
Text #18
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.1915. Min Act: -0.1700
Data Index: 853073 (C4 (Web Text))
Max Activating Token Index: 342
Click toggle to see full text
Truncated
Full Text #18
Text #19
Max Range: 4.6603. Min Range: -4.6603
Max Act: 4.1915. Min Act: -0.1699
Data Index: 1380133 (Python Code)
Max Activating Token Index: 104
Click toggle to see full text
Truncated
Full Text #19