Model: GELU Model: 4 Layers, 2048 Neurons per Layer
Dataset: 80% C4 (Web Text) and 20% Python Code
Neuron 1984 in Layer 0
Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-4l')
Text #0
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.5671. Min Act: -0.1700
Data Index: 357768 (C4 (Web Text))
Max Activating Token Index: 902
Click toggle to see full text
Truncated
Full Text #0
Text #1
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.5186. Min Act: -0.1700
Data Index: 295769 (C4 (Web Text))
Max Activating Token Index: 767
Click toggle to see full text
Truncated
Full Text #1
Text #2
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.3559. Min Act: -0.1700
Data Index: 100583 (C4 (Web Text))
Max Activating Token Index: 615
Click toggle to see full text
Truncated
Full Text #2
Text #3
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.3559. Min Act: -0.1700
Data Index: 194794 (C4 (Web Text))
Max Activating Token Index: 300
Click toggle to see full text
Truncated
Full Text #3
Text #4
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.3723. Min Act: -0.1700
Data Index: 386470 (C4 (Web Text))
Max Activating Token Index: 56
Click toggle to see full text
Truncated
Full Text #4
Text #5
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.3231. Min Act: -0.1700
Data Index: 1237298 (C4 (Web Text))
Max Activating Token Index: 34
Click toggle to see full text
Truncated
Full Text #5
Text #6
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.3395. Min Act: -0.1700
Data Index: 108170 (C4 (Web Text))
Max Activating Token Index: 947
Click toggle to see full text
Truncated
Full Text #6
Text #7
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.3067. Min Act: -0.1700
Data Index: 1538791 (Python Code)
Max Activating Token Index: 651
Click toggle to see full text
Truncated
Full Text #7
Text #8
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.3231. Min Act: -0.1700
Data Index: 1016184 (C4 (Web Text))
Max Activating Token Index: 984
Click toggle to see full text
Truncated
Full Text #8
Text #9
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.2903. Min Act: -0.1700
Data Index: 838100 (C4 (Web Text))
Max Activating Token Index: 969
Click toggle to see full text
Truncated
Full Text #9
Text #10
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.2903. Min Act: -0.1700
Data Index: 1510003 (Python Code)
Max Activating Token Index: 277
Click toggle to see full text
Truncated
Full Text #10
Text #11
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.2903. Min Act: -0.1700
Data Index: 1449715 (Python Code)
Max Activating Token Index: 889
Click toggle to see full text
Truncated
Full Text #11
Text #12
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.2738. Min Act: -0.1700
Data Index: 891386 (C4 (Web Text))
Max Activating Token Index: 14
Click toggle to see full text
Truncated
Full Text #12
Text #13
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.2738. Min Act: -0.1700
Data Index: 1425278 (Python Code)
Max Activating Token Index: 749
Click toggle to see full text
Truncated
Full Text #13
Text #14
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.2573. Min Act: -0.1700
Data Index: 1401542 (Python Code)
Max Activating Token Index: 366
Click toggle to see full text
Truncated
Full Text #14
Text #15
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.2573. Min Act: -0.1700
Data Index: 1434256 (Python Code)
Max Activating Token Index: 640
Click toggle to see full text
Truncated
Full Text #15
Text #16
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.2573. Min Act: -0.1700
Data Index: 449378 (C4 (Web Text))
Max Activating Token Index: 21
Click toggle to see full text
Truncated
Full Text #16
Text #17
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.2573. Min Act: -0.1700
Data Index: 397689 (C4 (Web Text))
Max Activating Token Index: 356
Click toggle to see full text
Truncated
Full Text #17
Text #18
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.2243. Min Act: -0.1700
Data Index: 1545688 (Python Code)
Max Activating Token Index: 799
Click toggle to see full text
Truncated
Full Text #18
Text #19
Max Range: 2.5671. Min Range: -2.5671
Max Act: 2.2243. Min Act: -0.1700
Data Index: 1552026 (Python Code)
Max Activating Token Index: 295
Click toggle to see full text
Truncated
Full Text #19