Model: GELU Model: 3 Layers, 2048 Neurons per Layer
Dataset: 80% C4 (Web Text) and 20% Python Code
Neuron 2025 in Layer 0
Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-3l')
Text #0
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.5718. Min Act: -0.1700
Data Index: 1394849 (Python Code)
Max Activating Token Index: 910
Click toggle to see full text
Truncated
Full Text #0
Text #1
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.5557. Min Act: -0.1700
Data Index: 13124 (C4 (Web Text))
Max Activating Token Index: 846
Click toggle to see full text
Truncated
Full Text #1
Text #2
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.5395. Min Act: -0.1700
Data Index: 1544855 (Python Code)
Max Activating Token Index: 317
Click toggle to see full text
Truncated
Full Text #2
Text #3
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4910. Min Act: -0.1700
Data Index: 904670 (C4 (Web Text))
Max Activating Token Index: 731
Click toggle to see full text
Truncated
Full Text #3
Text #4
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4748. Min Act: -0.1700
Data Index: 452465 (C4 (Web Text))
Max Activating Token Index: 419
Click toggle to see full text
Truncated
Full Text #4
Text #5
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4586. Min Act: -0.1700
Data Index: 1624396 (Python Code)
Max Activating Token Index: 186
Click toggle to see full text
Truncated
Full Text #5
Text #6
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4586. Min Act: -0.1700
Data Index: 1364079 (Python Code)
Max Activating Token Index: 455
Click toggle to see full text
Truncated
Full Text #6
Text #7
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4586. Min Act: -0.1700
Data Index: 877800 (C4 (Web Text))
Max Activating Token Index: 705
Click toggle to see full text
Truncated
Full Text #7
Text #8
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4260. Min Act: -0.1700
Data Index: 1350837 (C4 (Web Text))
Max Activating Token Index: 52
Click toggle to see full text
Truncated
Full Text #8
Text #9
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4586. Min Act: -0.1700
Data Index: 2510 (C4 (Web Text))
Max Activating Token Index: 585
Click toggle to see full text
Truncated
Full Text #9
Text #10
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4423. Min Act: -0.1700
Data Index: 16431 (C4 (Web Text))
Max Activating Token Index: 595
Click toggle to see full text
Truncated
Full Text #10
Text #11
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4260. Min Act: -0.1700
Data Index: 1048261 (C4 (Web Text))
Max Activating Token Index: 631
Click toggle to see full text
Truncated
Full Text #11
Text #12
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4260. Min Act: -0.1700
Data Index: 711142 (C4 (Web Text))
Max Activating Token Index: 381
Click toggle to see full text
Truncated
Full Text #12
Text #13
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4260. Min Act: -0.1700
Data Index: 229007 (C4 (Web Text))
Max Activating Token Index: 795
Click toggle to see full text
Truncated
Full Text #13
Text #14
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4260. Min Act: -0.1700
Data Index: 464108 (C4 (Web Text))
Max Activating Token Index: 551
Click toggle to see full text
Truncated
Full Text #14
Text #15
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4260. Min Act: -0.1700
Data Index: 517491 (C4 (Web Text))
Max Activating Token Index: 147
Click toggle to see full text
Truncated
Full Text #15
Text #16
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4097. Min Act: -0.1700
Data Index: 1186515 (C4 (Web Text))
Max Activating Token Index: 478
Click toggle to see full text
Truncated
Full Text #16
Text #17
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4260. Min Act: -0.1700
Data Index: 1192371 (C4 (Web Text))
Max Activating Token Index: 654
Click toggle to see full text
Truncated
Full Text #17
Text #18
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4097. Min Act: -0.1700
Data Index: 1592590 (Python Code)
Max Activating Token Index: 681
Click toggle to see full text
Truncated
Full Text #18
Text #19
Max Range: 2.5718. Min Range: -2.5718
Max Act: 2.4097. Min Act: -0.1700
Data Index: 1412826 (Python Code)
Max Activating Token Index: 965
Click toggle to see full text
Truncated
Full Text #19