Model: GELU Model: 3 Layers, 2048 Neurons per Layer
Dataset: 80% C4 (Web Text) and 20% Python Code
Neuron 3 in Layer 2
Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-3l')
Text #0
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.9932. Min Act: -0.1700
Data Index: 378616 (C4 (Web Text))
Max Activating Token Index: 133
Click toggle to see full text
Truncated
Full Text #0
Text #1
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.7119. Min Act: -0.1700
Data Index: 1106621 (C4 (Web Text))
Max Activating Token Index: 643
Click toggle to see full text
Truncated
Full Text #1
Text #2
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.7119. Min Act: -0.1700
Data Index: 894854 (C4 (Web Text))
Max Activating Token Index: 364
Click toggle to see full text
Truncated
Full Text #2
Text #3
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.7119. Min Act: -0.1700
Data Index: 831307 (C4 (Web Text))
Max Activating Token Index: 749
Click toggle to see full text
Truncated
Full Text #3
Text #4
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.7119. Min Act: -0.1700
Data Index: 939423 (C4 (Web Text))
Max Activating Token Index: 916
Click toggle to see full text
Truncated
Full Text #4
Text #5
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.6807. Min Act: -0.1700
Data Index: 731186 (C4 (Web Text))
Max Activating Token Index: 40
Click toggle to see full text
Truncated
Full Text #5
Text #6
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.6494. Min Act: -0.1700
Data Index: 129535 (C4 (Web Text))
Max Activating Token Index: 13
Click toggle to see full text
Truncated
Full Text #6
Text #7
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.6182. Min Act: -0.1700
Data Index: 268307 (C4 (Web Text))
Max Activating Token Index: 104
Click toggle to see full text
Truncated
Full Text #7
Text #8
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.5869. Min Act: -0.1700
Data Index: 984325 (C4 (Web Text))
Max Activating Token Index: 76
Click toggle to see full text
Truncated
Full Text #8
Text #9
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.6182. Min Act: -0.1700
Data Index: 61 (C4 (Web Text))
Max Activating Token Index: 867
Click toggle to see full text
Truncated
Full Text #9
Text #10
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.5557. Min Act: -0.1700
Data Index: 537908 (C4 (Web Text))
Max Activating Token Index: 44
Click toggle to see full text
Truncated
Full Text #10
Text #11
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.5557. Min Act: -0.1700
Data Index: 1128388 (C4 (Web Text))
Max Activating Token Index: 57
Click toggle to see full text
Truncated
Full Text #11
Text #12
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.5244. Min Act: -0.1700
Data Index: 300448 (C4 (Web Text))
Max Activating Token Index: 18
Click toggle to see full text
Truncated
Full Text #12
Text #13
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.5557. Min Act: -0.1700
Data Index: 234004 (C4 (Web Text))
Max Activating Token Index: 16
Click toggle to see full text
Truncated
Full Text #13
Text #14
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.4932. Min Act: -0.1700
Data Index: 324145 (C4 (Web Text))
Max Activating Token Index: 814
Click toggle to see full text
Truncated
Full Text #14
Text #15
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.4619. Min Act: -0.1700
Data Index: 1059111 (C4 (Web Text))
Max Activating Token Index: 718
Click toggle to see full text
Truncated
Full Text #15
Text #16
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.4619. Min Act: -0.1700
Data Index: 813435 (C4 (Web Text))
Max Activating Token Index: 915
Click toggle to see full text
Truncated
Full Text #16
Text #17
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.4619. Min Act: -0.1700
Data Index: 96296 (C4 (Web Text))
Max Activating Token Index: 592
Click toggle to see full text
Truncated
Full Text #17
Text #18
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.4619. Min Act: -0.1700
Data Index: 1082542 (C4 (Web Text))
Max Activating Token Index: 152
Click toggle to see full text
Truncated
Full Text #18
Text #19
Max Range: 4.9932. Min Range: -4.9932
Max Act: 4.3994. Min Act: -0.1700
Data Index: 898578 (C4 (Web Text))
Max Activating Token Index: 281
Click toggle to see full text
Truncated
Full Text #19