Model: GELU Model: 4 Layers, 2048 Neurons per Layer
Dataset: 80% C4 (Web Text) and 20% Python Code
Neuron 1999 in Layer 2
Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-4l')
Text #0
Max Range: 5.4166. Min Range: -5.4166
Max Act: 5.4166. Min Act: -0.1700
Data Index: 900180 (C4 (Web Text))
Max Activating Token Index: 1023
Click toggle to see full text
Truncated
Full Text #0
Text #1
Max Range: 5.4166. Min Range: -5.4166
Max Act: 5.1354. Min Act: -0.1700
Data Index: 998497 (C4 (Web Text))
Max Activating Token Index: 825
Click toggle to see full text
Truncated
Full Text #1
Text #2
Max Range: 5.4166. Min Range: -5.4166
Max Act: 5.1354. Min Act: -0.1700
Data Index: 1225971 (C4 (Web Text))
Max Activating Token Index: 1023
Click toggle to see full text
Truncated
Full Text #2
Text #3
Max Range: 5.4166. Min Range: -5.4166
Max Act: 5.1041. Min Act: -0.1700
Data Index: 1066047 (C4 (Web Text))
Max Activating Token Index: 1023
Click toggle to see full text
Truncated
Full Text #3
Text #4
Max Range: 5.4166. Min Range: -5.4166
Max Act: 5.0416. Min Act: -0.1700
Data Index: 738169 (C4 (Web Text))
Max Activating Token Index: 299
Click toggle to see full text
Truncated
Full Text #4
Text #5
Max Range: 5.4166. Min Range: -5.4166
Max Act: 5.0416. Min Act: -0.1700
Data Index: 1070290 (C4 (Web Text))
Max Activating Token Index: 745
Click toggle to see full text
Truncated
Full Text #5
Text #6
Max Range: 5.4166. Min Range: -5.4166
Max Act: 5.0729. Min Act: -0.1700
Data Index: 253075 (C4 (Web Text))
Max Activating Token Index: 1023
Click toggle to see full text
Truncated
Full Text #6
Text #7
Max Range: 5.4166. Min Range: -5.4166
Max Act: 5.0104. Min Act: -0.1700
Data Index: 444925 (C4 (Web Text))
Max Activating Token Index: 1023
Click toggle to see full text
Truncated
Full Text #7
Text #8
Max Range: 5.4166. Min Range: -5.4166
Max Act: 4.9791. Min Act: -0.1700
Data Index: 1144536 (C4 (Web Text))
Max Activating Token Index: 301
Click toggle to see full text
Truncated
Full Text #8
Text #9
Max Range: 5.4166. Min Range: -5.4166
Max Act: 4.9479. Min Act: -0.1700
Data Index: 797994 (C4 (Web Text))
Max Activating Token Index: 1023
Click toggle to see full text
Truncated
Full Text #9
Text #10
Max Range: 5.4166. Min Range: -5.4166
Max Act: 4.9166. Min Act: -0.1700
Data Index: 130460 (C4 (Web Text))
Max Activating Token Index: 1023
Click toggle to see full text
Truncated
Full Text #10
Text #11
Max Range: 5.4166. Min Range: -5.4166
Max Act: 4.8854. Min Act: -0.1700
Data Index: 171091 (C4 (Web Text))
Max Activating Token Index: 1023
Click toggle to see full text
Truncated
Full Text #11
Text #12
Max Range: 5.4166. Min Range: -5.4166
Max Act: 4.8854. Min Act: -0.1700
Data Index: 553766 (C4 (Web Text))
Max Activating Token Index: 1023
Click toggle to see full text
Truncated
Full Text #12
Text #13
Max Range: 5.4166. Min Range: -5.4166
Max Act: 4.8541. Min Act: -0.1700
Data Index: 596787 (C4 (Web Text))
Max Activating Token Index: 1023
Click toggle to see full text
Truncated
Full Text #13
Text #14
Max Range: 5.4166. Min Range: -5.4166
Max Act: 4.9166. Min Act: -0.1700
Data Index: 821085 (C4 (Web Text))
Max Activating Token Index: 882
Click toggle to see full text
Truncated
Full Text #14
Text #15
Max Range: 5.4166. Min Range: -5.4166
Max Act: 4.9166. Min Act: -0.1700
Data Index: 402465 (C4 (Web Text))
Max Activating Token Index: 1023
Click toggle to see full text
Truncated
Full Text #15
Text #16
Max Range: 5.4166. Min Range: -5.4166
Max Act: 4.8854. Min Act: -0.1700
Data Index: 1251402 (C4 (Web Text))
Max Activating Token Index: 571
Click toggle to see full text
Truncated
Full Text #16
Text #17
Max Range: 5.4166. Min Range: -5.4166
Max Act: 4.8854. Min Act: -0.1700
Data Index: 1069579 (C4 (Web Text))
Max Activating Token Index: 1023
Click toggle to see full text
Truncated
Full Text #17
Text #18
Max Range: 5.4166. Min Range: -5.4166
Max Act: 4.8541. Min Act: -0.1700
Data Index: 329485 (C4 (Web Text))
Max Activating Token Index: 1023
Click toggle to see full text
Truncated
Full Text #18
Text #19
Max Range: 5.4166. Min Range: -5.4166
Max Act: 4.8854. Min Act: -0.1700
Data Index: 1174826 (C4 (Web Text))
Max Activating Token Index: 733
Click toggle to see full text
Truncated
Full Text #19