Model: GELU Model: 4 Layers, 2048 Neurons per Layer
Dataset: 80% C4 (Web Text) and 20% Python Code
Neuron 1273 in Layer 3
Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-4l')
Text #0
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.9696. Min Act: -0.1700
Data Index: 196385 (C4 (Web Text))
Max Activating Token Index: 440
Click toggle to see full text
Truncated
Full Text #0
Text #1
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.9384. Min Act: -0.1700
Data Index: 678211 (C4 (Web Text))
Max Activating Token Index: 391
Click toggle to see full text
Truncated
Full Text #1
Text #2
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.9227. Min Act: -0.1700
Data Index: 518474 (C4 (Web Text))
Max Activating Token Index: 110
Click toggle to see full text
Truncated
Full Text #2
Text #3
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.8290. Min Act: -0.1700
Data Index: 1469478 (Python Code)
Max Activating Token Index: 412
Click toggle to see full text
Truncated
Full Text #3
Text #4
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.9227. Min Act: -0.1700
Data Index: 1455472 (Python Code)
Max Activating Token Index: 598
Click toggle to see full text
Truncated
Full Text #4
Text #5
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.7352. Min Act: -0.1700
Data Index: 418690 (C4 (Web Text))
Max Activating Token Index: 540
Click toggle to see full text
Truncated
Full Text #5
Text #6
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.7196. Min Act: -0.1700
Data Index: 1650077 (Python Code)
Max Activating Token Index: 579
Click toggle to see full text
Truncated
Full Text #6
Text #7
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.6727. Min Act: -0.1700
Data Index: 614540 (C4 (Web Text))
Max Activating Token Index: 811
Click toggle to see full text
Truncated
Full Text #7
Text #8
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.7040. Min Act: -0.1700
Data Index: 1384207 (Python Code)
Max Activating Token Index: 923
Click toggle to see full text
Truncated
Full Text #8
Text #9
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.6259. Min Act: -0.1700
Data Index: 59147 (C4 (Web Text))
Max Activating Token Index: 379
Click toggle to see full text
Truncated
Full Text #9
Text #10
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.6102. Min Act: -0.1700
Data Index: 1614589 (Python Code)
Max Activating Token Index: 155
Click toggle to see full text
Truncated
Full Text #10
Text #11
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.5321. Min Act: -0.1700
Data Index: 1393228 (Python Code)
Max Activating Token Index: 673
Click toggle to see full text
Truncated
Full Text #11
Text #12
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.5477. Min Act: -0.1700
Data Index: 650364 (C4 (Web Text))
Max Activating Token Index: 615
Click toggle to see full text
Truncated
Full Text #12
Text #13
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.6259. Min Act: -0.1700
Data Index: 1488861 (Python Code)
Max Activating Token Index: 1000
Click toggle to see full text
Truncated
Full Text #13
Text #14
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.6102. Min Act: -0.1700
Data Index: 1381421 (Python Code)
Max Activating Token Index: 486
Click toggle to see full text
Truncated
Full Text #14
Text #15
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.6102. Min Act: -0.1700
Data Index: 1520827 (Python Code)
Max Activating Token Index: 604
Click toggle to see full text
Truncated
Full Text #15
Text #16
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.6102. Min Act: -0.1700
Data Index: 1095857 (C4 (Web Text))
Max Activating Token Index: 585
Click toggle to see full text
Truncated
Full Text #16
Text #17
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.5165. Min Act: -0.1700
Data Index: 60768 (C4 (Web Text))
Max Activating Token Index: 894
Click toggle to see full text
Truncated
Full Text #17
Text #18
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.4696. Min Act: -0.1700
Data Index: 1443537 (Python Code)
Max Activating Token Index: 743
Click toggle to see full text
Truncated
Full Text #18
Text #19
Max Range: 6.9696. Min Range: -6.9696
Max Act: 6.4540. Min Act: -0.1700
Data Index: 1592465 (Python Code)
Max Activating Token Index: 855
Click toggle to see full text
Truncated
Full Text #19