Model: GELU Model: 4 Layers, 2048 Neurons per Layer
Dataset: 80% C4 (Web Text) and 20% Python Code
Neuron 1603 in Layer 2
Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-4l')
Text #0
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.7901. Min Act: -0.1700
Data Index: 449190 (C4 (Web Text))
Max Activating Token Index: 963
Click toggle to see full text
Truncated
Full Text #0
Text #1
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.7276. Min Act: -0.1700
Data Index: 325695 (C4 (Web Text))
Max Activating Token Index: 561
Click toggle to see full text
Truncated
Full Text #1
Text #2
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.7276. Min Act: -0.1700
Data Index: 443675 (C4 (Web Text))
Max Activating Token Index: 84
Click toggle to see full text
Truncated
Full Text #2
Text #3
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.5401. Min Act: -0.1700
Data Index: 338177 (C4 (Web Text))
Max Activating Token Index: 776
Click toggle to see full text
Truncated
Full Text #3
Text #4
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.6338. Min Act: -0.1700
Data Index: 1093795 (C4 (Web Text))
Max Activating Token Index: 476
Click toggle to see full text
Truncated
Full Text #4
Text #5
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.5713. Min Act: -0.1700
Data Index: 944485 (C4 (Web Text))
Max Activating Token Index: 816
Click toggle to see full text
Truncated
Full Text #5
Text #6
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.5713. Min Act: -0.1700
Data Index: 1434076 (Python Code)
Max Activating Token Index: 578
Click toggle to see full text
Truncated
Full Text #6
Text #7
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.5713. Min Act: -0.1700
Data Index: 16536 (C4 (Web Text))
Max Activating Token Index: 583
Click toggle to see full text
Truncated
Full Text #7
Text #8
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.6026. Min Act: -0.1700
Data Index: 435262 (C4 (Web Text))
Max Activating Token Index: 377
Click toggle to see full text
Truncated
Full Text #8
Text #9
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.5713. Min Act: -0.1700
Data Index: 237107 (C4 (Web Text))
Max Activating Token Index: 725
Click toggle to see full text
Truncated
Full Text #9
Text #10
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.5713. Min Act: -0.1700
Data Index: 5350 (C4 (Web Text))
Max Activating Token Index: 359
Click toggle to see full text
Truncated
Full Text #10
Text #11
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.5401. Min Act: -0.1700
Data Index: 1271527 (C4 (Web Text))
Max Activating Token Index: 857
Click toggle to see full text
Truncated
Full Text #11
Text #12
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.5401. Min Act: -0.1700
Data Index: 858512 (C4 (Web Text))
Max Activating Token Index: 708
Click toggle to see full text
Truncated
Full Text #12
Text #13
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.4463. Min Act: -0.1700
Data Index: 1413119 (Python Code)
Max Activating Token Index: 586
Click toggle to see full text
Truncated
Full Text #13
Text #14
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.5088. Min Act: -0.1700
Data Index: 1056129 (C4 (Web Text))
Max Activating Token Index: 703
Click toggle to see full text
Truncated
Full Text #14
Text #15
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.5088. Min Act: -0.1700
Data Index: 129505 (C4 (Web Text))
Max Activating Token Index: 839
Click toggle to see full text
Truncated
Full Text #15
Text #16
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.4776. Min Act: -0.1700
Data Index: 543236 (C4 (Web Text))
Max Activating Token Index: 677
Click toggle to see full text
Truncated
Full Text #16
Text #17
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.5088. Min Act: -0.1700
Data Index: 1518889 (Python Code)
Max Activating Token Index: 552
Click toggle to see full text
Truncated
Full Text #17
Text #18
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.4776. Min Act: -0.1700
Data Index: 1142852 (C4 (Web Text))
Max Activating Token Index: 861
Click toggle to see full text
Truncated
Full Text #18
Text #19
Max Range: 4.7901. Min Range: -4.7901
Max Act: 4.4776. Min Act: -0.1700
Data Index: 1523588 (Python Code)
Max Activating Token Index: 505
Click toggle to see full text
Truncated
Full Text #19