Model: GELU Model: 3 Layers, 2048 Neurons per Layer
Dataset: 80% C4 (Web Text) and 20% Python Code
Neuron 2011 in Layer 1
Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-3l')
Text #0
Max Range: 3.1307. Min Range: -3.1307
Max Act: 3.1307. Min Act: -0.1700
Data Index: 617092 (C4 (Web Text))
Max Activating Token Index: 585
Click toggle to see full text
Truncated
Full Text #0
Text #1
Max Range: 3.1307. Min Range: -3.1307
Max Act: 3.0676. Min Act: -0.1700
Data Index: 877091 (C4 (Web Text))
Max Activating Token Index: 100
Click toggle to see full text
Truncated
Full Text #1
Text #2
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.8936. Min Act: -0.1700
Data Index: 1320761 (C4 (Web Text))
Max Activating Token Index: 680
Click toggle to see full text
Truncated
Full Text #2
Text #3
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.8618. Min Act: -0.1700
Data Index: 922164 (C4 (Web Text))
Max Activating Token Index: 917
Click toggle to see full text
Truncated
Full Text #3
Text #4
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.8777. Min Act: -0.1700
Data Index: 1119023 (C4 (Web Text))
Max Activating Token Index: 841
Click toggle to see full text
Truncated
Full Text #4
Text #5
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.8777. Min Act: -0.1700
Data Index: 810880 (C4 (Web Text))
Max Activating Token Index: 243
Click toggle to see full text
Truncated
Full Text #5
Text #6
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.8460. Min Act: -0.1700
Data Index: 785814 (C4 (Web Text))
Max Activating Token Index: 121
Click toggle to see full text
Truncated
Full Text #6
Text #7
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.8301. Min Act: -0.1700
Data Index: 456350 (C4 (Web Text))
Max Activating Token Index: 297
Click toggle to see full text
Truncated
Full Text #7
Text #8
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.8301. Min Act: -0.1700
Data Index: 769032 (C4 (Web Text))
Max Activating Token Index: 834
Click toggle to see full text
Truncated
Full Text #8
Text #9
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.7663. Min Act: -0.1700
Data Index: 917963 (C4 (Web Text))
Max Activating Token Index: 9
Click toggle to see full text
Truncated
Full Text #9
Text #10
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.7823. Min Act: -0.1700
Data Index: 274760 (C4 (Web Text))
Max Activating Token Index: 360
Click toggle to see full text
Truncated
Full Text #10
Text #11
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.7982. Min Act: -0.1700
Data Index: 1135113 (C4 (Web Text))
Max Activating Token Index: 985
Click toggle to see full text
Truncated
Full Text #11
Text #12
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.7663. Min Act: -0.1700
Data Index: 68231 (C4 (Web Text))
Max Activating Token Index: 698
Click toggle to see full text
Truncated
Full Text #12
Text #13
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.7504. Min Act: -0.1700
Data Index: 1326193 (C4 (Web Text))
Max Activating Token Index: 752
Click toggle to see full text
Truncated
Full Text #13
Text #14
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.7982. Min Act: -0.1700
Data Index: 762208 (C4 (Web Text))
Max Activating Token Index: 11
Click toggle to see full text
Truncated
Full Text #14
Text #15
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.7982. Min Act: -0.1700
Data Index: 1278828 (C4 (Web Text))
Max Activating Token Index: 21
Click toggle to see full text
Truncated
Full Text #15
Text #16
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.7663. Min Act: -0.1700
Data Index: 1486476 (Python Code)
Max Activating Token Index: 149
Click toggle to see full text
Truncated
Full Text #16
Text #17
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.7663. Min Act: -0.1700
Data Index: 1398155 (Python Code)
Max Activating Token Index: 835
Click toggle to see full text
Truncated
Full Text #17
Text #18
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.7504. Min Act: -0.1700
Data Index: 661893 (C4 (Web Text))
Max Activating Token Index: 998
Click toggle to see full text
Truncated
Full Text #18
Text #19
Max Range: 3.1307. Min Range: -3.1307
Max Act: 2.7344. Min Act: -0.1700
Data Index: 684130 (C4 (Web Text))
Max Activating Token Index: 882
Click toggle to see full text
Truncated
Full Text #19