Model: GELU Model: 2 Layers, 2048 Neurons per Layer
Dataset: 80% C4 (Web Text) and 20% Python Code
Neuron 989 in Layer 1
Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-2l')
Text #0
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.2423. Min Act: -0.1700
Data Index: 771533 (C4 (Web Text))
Max Activating Token Index: 750
Click toggle to see full text
Truncated
Full Text #0
Text #1
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.2736. Min Act: -0.1700
Data Index: 1416626 (Python Code)
Max Activating Token Index: 384
Click toggle to see full text
Truncated
Full Text #1
Text #2
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.2423. Min Act: -0.1700
Data Index: 410167 (C4 (Web Text))
Max Activating Token Index: 90
Click toggle to see full text
Truncated
Full Text #2
Text #3
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.2423. Min Act: -0.1700
Data Index: 794907 (C4 (Web Text))
Max Activating Token Index: 793
Click toggle to see full text
Truncated
Full Text #3
Text #4
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.1329. Min Act: -0.1700
Data Index: 387051 (C4 (Web Text))
Max Activating Token Index: 664
Click toggle to see full text
Truncated
Full Text #4
Text #5
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.0861. Min Act: -0.1700
Data Index: 1524359 (Python Code)
Max Activating Token Index: 655
Click toggle to see full text
Truncated
Full Text #5
Text #6
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.0861. Min Act: -0.1700
Data Index: 1471772 (Python Code)
Max Activating Token Index: 360
Click toggle to see full text
Truncated
Full Text #6
Text #7
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.1017. Min Act: -0.1700
Data Index: 588970 (C4 (Web Text))
Max Activating Token Index: 486
Click toggle to see full text
Truncated
Full Text #7
Text #8
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.1486. Min Act: -0.1700
Data Index: 1518047 (Python Code)
Max Activating Token Index: 805
Click toggle to see full text
Truncated
Full Text #8
Text #9
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.0861. Min Act: -0.1700
Data Index: 178043 (C4 (Web Text))
Max Activating Token Index: 41
Click toggle to see full text
Truncated
Full Text #9
Text #10
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.1017. Min Act: -0.1700
Data Index: 1118319 (C4 (Web Text))
Max Activating Token Index: 129
Click toggle to see full text
Truncated
Full Text #10
Text #11
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.1017. Min Act: -0.1700
Data Index: 929329 (C4 (Web Text))
Max Activating Token Index: 907
Click toggle to see full text
Truncated
Full Text #11
Text #12
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.1017. Min Act: -0.1700
Data Index: 1416260 (Python Code)
Max Activating Token Index: 197
Click toggle to see full text
Truncated
Full Text #12
Text #13
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.0704. Min Act: -0.1700
Data Index: 422238 (C4 (Web Text))
Max Activating Token Index: 513
Click toggle to see full text
Truncated
Full Text #13
Text #14
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.0861. Min Act: -0.1700
Data Index: 1498804 (Python Code)
Max Activating Token Index: 232
Click toggle to see full text
Truncated
Full Text #14
Text #15
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.0235. Min Act: -0.1700
Data Index: 674362 (C4 (Web Text))
Max Activating Token Index: 581
Click toggle to see full text
Truncated
Full Text #15
Text #16
Max Range: 4.2736. Min Range: -4.2736
Max Act: 3.9923. Min Act: -0.1700
Data Index: 669373 (C4 (Web Text))
Max Activating Token Index: 44
Click toggle to see full text
Truncated
Full Text #16
Text #17
Max Range: 4.2736. Min Range: -4.2736
Max Act: 3.9297. Min Act: -0.1700
Data Index: 1394393 (Python Code)
Max Activating Token Index: 966
Click toggle to see full text
Truncated
Full Text #17
Text #18
Max Range: 4.2736. Min Range: -4.2736
Max Act: 4.0079. Min Act: -0.1700
Data Index: 598648 (C4 (Web Text))
Max Activating Token Index: 928
Click toggle to see full text
Truncated
Full Text #18
Text #19
Max Range: 4.2736. Min Range: -4.2736
Max Act: 3.9766. Min Act: -0.1700
Data Index: 1445197 (Python Code)
Max Activating Token Index: 427
Click toggle to see full text
Truncated
Full Text #19