Model: GELU Model: 4 Layers, 2048 Neurons per Layer
Dataset: 80% C4 (Web Text) and 20% Python Code
Neuron 54 in Layer 0
Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-4l')
Text #0
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9444. Min Act: -0.1700
Data Index: 563051 (C4 (Web Text))
Max Activating Token Index: 551
Click toggle to see full text
Truncated
Full Text #0
Text #1
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9359. Min Act: -0.1700
Data Index: 1136065 (C4 (Web Text))
Max Activating Token Index: 556
Click toggle to see full text
Truncated
Full Text #1
Text #2
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9359. Min Act: -0.1700
Data Index: 1254304 (C4 (Web Text))
Max Activating Token Index: 548
Click toggle to see full text
Truncated
Full Text #2
Text #3
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9274. Min Act: -0.1700
Data Index: 366535 (C4 (Web Text))
Max Activating Token Index: 544
Click toggle to see full text
Truncated
Full Text #3
Text #4
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9274. Min Act: -0.1700
Data Index: 985099 (C4 (Web Text))
Max Activating Token Index: 545
Click toggle to see full text
Truncated
Full Text #4
Text #5
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9189. Min Act: -0.1700
Data Index: 1594282 (Python Code)
Max Activating Token Index: 544
Click toggle to see full text
Truncated
Full Text #5
Text #6
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9189. Min Act: -0.1700
Data Index: 314066 (C4 (Web Text))
Max Activating Token Index: 567
Click toggle to see full text
Truncated
Full Text #6
Text #7
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9104. Min Act: -0.1700
Data Index: 87067 (C4 (Web Text))
Max Activating Token Index: 566
Click toggle to see full text
Truncated
Full Text #7
Text #8
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9189. Min Act: -0.1700
Data Index: 587692 (C4 (Web Text))
Max Activating Token Index: 571
Click toggle to see full text
Truncated
Full Text #8
Text #9
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9189. Min Act: -0.1700
Data Index: 971389 (C4 (Web Text))
Max Activating Token Index: 539
Click toggle to see full text
Truncated
Full Text #9
Text #10
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9104. Min Act: -0.1700
Data Index: 428128 (C4 (Web Text))
Max Activating Token Index: 578
Click toggle to see full text
Truncated
Full Text #10
Text #11
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9104. Min Act: -0.1700
Data Index: 732740 (C4 (Web Text))
Max Activating Token Index: 561
Click toggle to see full text
Truncated
Full Text #11
Text #12
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9104. Min Act: -0.1700
Data Index: 85133 (C4 (Web Text))
Max Activating Token Index: 585
Click toggle to see full text
Truncated
Full Text #12
Text #13
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9019. Min Act: -0.1700
Data Index: 255995 (C4 (Web Text))
Max Activating Token Index: 555
Click toggle to see full text
Truncated
Full Text #13
Text #14
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9019. Min Act: -0.1700
Data Index: 114874 (C4 (Web Text))
Max Activating Token Index: 574
Click toggle to see full text
Truncated
Full Text #14
Text #15
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9019. Min Act: -0.1700
Data Index: 589247 (C4 (Web Text))
Max Activating Token Index: 590
Click toggle to see full text
Truncated
Full Text #15
Text #16
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9104. Min Act: -0.1700
Data Index: 1041001 (C4 (Web Text))
Max Activating Token Index: 563
Click toggle to see full text
Truncated
Full Text #16
Text #17
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9019. Min Act: -0.1700
Data Index: 325147 (C4 (Web Text))
Max Activating Token Index: 559
Click toggle to see full text
Truncated
Full Text #17
Text #18
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9019. Min Act: -0.1700
Data Index: 775136 (C4 (Web Text))
Max Activating Token Index: 570
Click toggle to see full text
Truncated
Full Text #18
Text #19
Max Range: 1.9444. Min Range: -1.9444
Max Act: 1.9019. Min Act: -0.1700
Data Index: 911831 (C4 (Web Text))
Max Activating Token Index: 548
Click toggle to see full text
Truncated
Full Text #19