Model: GELU Model: 3 Layers, 2048 Neurons per Layer
Dataset: 80% C4 (Web Text) and 20% Python Code
Neuron 79 in Layer 1
Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-3l')
Text #0
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.9986. Min Act: -0.1700
Data Index: 810167 (C4 (Web Text))
Max Activating Token Index: 394
Click toggle to see full text
Truncated
Full Text #0
Text #1
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.9035. Min Act: -0.1700
Data Index: 1129537 (C4 (Web Text))
Max Activating Token Index: 66
Click toggle to see full text
Truncated
Full Text #1
Text #2
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.9194. Min Act: -0.1700
Data Index: 465820 (C4 (Web Text))
Max Activating Token Index: 1023
Click toggle to see full text
Truncated
Full Text #2
Text #3
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.8877. Min Act: -0.1700
Data Index: 1598691 (Python Code)
Max Activating Token Index: 80
Click toggle to see full text
Truncated
Full Text #3
Text #4
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.8718. Min Act: -0.1700
Data Index: 987946 (C4 (Web Text))
Max Activating Token Index: 620
Click toggle to see full text
Truncated
Full Text #4
Text #5
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.8400. Min Act: -0.1700
Data Index: 441999 (C4 (Web Text))
Max Activating Token Index: 471
Click toggle to see full text
Truncated
Full Text #5
Text #6
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.8877. Min Act: -0.1700
Data Index: 469067 (C4 (Web Text))
Max Activating Token Index: 418
Click toggle to see full text
Truncated
Full Text #6
Text #7
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.8241. Min Act: -0.1700
Data Index: 1096408 (C4 (Web Text))
Max Activating Token Index: 888
Click toggle to see full text
Truncated
Full Text #7
Text #8
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.8559. Min Act: -0.1700
Data Index: 1596095 (Python Code)
Max Activating Token Index: 385
Click toggle to see full text
Truncated
Full Text #8
Text #9
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.7923. Min Act: -0.1700
Data Index: 1192832 (C4 (Web Text))
Max Activating Token Index: 85
Click toggle to see full text
Truncated
Full Text #9
Text #10
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.7923. Min Act: -0.1700
Data Index: 1011710 (C4 (Web Text))
Max Activating Token Index: 385
Click toggle to see full text
Truncated
Full Text #10
Text #11
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.7604. Min Act: -0.1700
Data Index: 1049233 (C4 (Web Text))
Max Activating Token Index: 12
Click toggle to see full text
Truncated
Full Text #11
Text #12
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.7923. Min Act: -0.1700
Data Index: 635507 (C4 (Web Text))
Max Activating Token Index: 816
Click toggle to see full text
Truncated
Full Text #12
Text #13
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.7763. Min Act: -0.1700
Data Index: 682393 (C4 (Web Text))
Max Activating Token Index: 292
Click toggle to see full text
Truncated
Full Text #13
Text #14
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.6965. Min Act: -0.1700
Data Index: 296037 (C4 (Web Text))
Max Activating Token Index: 810
Click toggle to see full text
Truncated
Full Text #14
Text #15
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.7923. Min Act: -0.1700
Data Index: 1290597 (C4 (Web Text))
Max Activating Token Index: 165
Click toggle to see full text
Truncated
Full Text #15
Text #16
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.7125. Min Act: -0.1700
Data Index: 30849 (C4 (Web Text))
Max Activating Token Index: 184
Click toggle to see full text
Truncated
Full Text #16
Text #17
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.7285. Min Act: -0.1700
Data Index: 739786 (C4 (Web Text))
Max Activating Token Index: 1023
Click toggle to see full text
Truncated
Full Text #17
Text #18
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.7604. Min Act: -0.1700
Data Index: 1563018 (Python Code)
Max Activating Token Index: 265
Click toggle to see full text
Truncated
Full Text #18
Text #19
Max Range: 2.9986. Min Range: -2.9986
Max Act: 2.7285. Min Act: -0.1700
Data Index: 1581578 (Python Code)
Max Activating Token Index: 492
Click toggle to see full text
Truncated
Full Text #19