Model: SoLU Model: 12 Layers, 6144 Neurons per Layer
Dataset: 80% C4 (Web Text) and 20% Python Code
Neuron 3650 in Layer 5
Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')
Text #0
Max Range: 1.9358. Min Range: -1.9358
Max Act: 1.9358. Min Act: -0.0001
Data Index: 562951 (C4 (Web Text))
Max Activating Token Index: 63
Click toggle to see full text
Truncated
Full Text #0
Text #1
Max Range: 1.9358. Min Range: -1.9358
Max Act: 1.4072. Min Act: -0.0001
Data Index: 634901 (C4 (Web Text))
Max Activating Token Index: 807
Click toggle to see full text
Truncated
Full Text #1
Text #2
Max Range: 1.9358. Min Range: -1.9358
Max Act: 1.4303. Min Act: -0.0001
Data Index: 835697 (C4 (Web Text))
Max Activating Token Index: 707
Click toggle to see full text
Truncated
Full Text #2
Text #3
Max Range: 1.9358. Min Range: -1.9358
Max Act: 1.4118. Min Act: -0.0001
Data Index: 295020 (C4 (Web Text))
Max Activating Token Index: 530
Click toggle to see full text
Truncated
Full Text #3
Text #4
Max Range: 1.9358. Min Range: -1.9358
Max Act: 1.2021. Min Act: -0.0001
Data Index: 766125 (C4 (Web Text))
Max Activating Token Index: 812
Click toggle to see full text
Truncated
Full Text #4
Text #5
Max Range: 1.9358. Min Range: -1.9358
Max Act: 1.2659. Min Act: -0.0001
Data Index: 458147 (C4 (Web Text))
Max Activating Token Index: 69
Click toggle to see full text
Truncated
Full Text #5
Text #6
Max Range: 1.9358. Min Range: -1.9358
Max Act: 1.2451. Min Act: -0.0001
Data Index: 967570 (C4 (Web Text))
Max Activating Token Index: 125
Click toggle to see full text
Truncated
Full Text #6
Text #7
Max Range: 1.9358. Min Range: -1.9358
Max Act: 1.2726. Min Act: -0.0001
Data Index: 58517 (C4 (Web Text))
Max Activating Token Index: 423
Click toggle to see full text
Truncated
Full Text #7
Text #8
Max Range: 1.9358. Min Range: -1.9358
Max Act: 1.1663. Min Act: -0.0001
Data Index: 371588 (C4 (Web Text))
Max Activating Token Index: 40
Click toggle to see full text
Truncated
Full Text #8
Text #9
Max Range: 1.9358. Min Range: -1.9358
Max Act: 1.2025. Min Act: -0.0001
Data Index: 630719 (C4 (Web Text))
Max Activating Token Index: 394
Click toggle to see full text
Truncated
Full Text #9
Text #10
Max Range: 1.9358. Min Range: -1.9358
Max Act: 1.1143. Min Act: -0.0001
Data Index: 766930 (C4 (Web Text))
Max Activating Token Index: 345
Click toggle to see full text
Truncated
Full Text #10
Text #11
Max Range: 1.9358. Min Range: -1.9358
Max Act: 1.0580. Min Act: -0.0001
Data Index: 1311108 (C4 (Web Text))
Max Activating Token Index: 308
Click toggle to see full text
Truncated
Full Text #11
Text #12
Max Range: 1.9358. Min Range: -1.9358
Max Act: 1.0558. Min Act: -0.0001
Data Index: 204254 (C4 (Web Text))
Max Activating Token Index: 426
Click toggle to see full text
Truncated
Full Text #12
Text #13
Max Range: 1.9358. Min Range: -1.9358
Max Act: 1.0228. Min Act: -0.0001
Data Index: 860563 (C4 (Web Text))
Max Activating Token Index: 4
Click toggle to see full text
Truncated
Full Text #13
Text #14
Max Range: 1.9358. Min Range: -1.9358
Max Act: 0.9839. Min Act: -0.0001
Data Index: 383129 (C4 (Web Text))
Max Activating Token Index: 107
Click toggle to see full text
Truncated
Full Text #14
Text #15
Max Range: 1.9358. Min Range: -1.9358
Max Act: 1.0136. Min Act: -0.0001
Data Index: 899573 (C4 (Web Text))
Max Activating Token Index: 965
Click toggle to see full text
Truncated
Full Text #15
Text #16
Max Range: 1.9358. Min Range: -1.9358
Max Act: 0.9377. Min Act: -0.0001
Data Index: 892955 (C4 (Web Text))
Max Activating Token Index: 362
Click toggle to see full text
Truncated
Full Text #16
Text #17
Max Range: 1.9358. Min Range: -1.9358
Max Act: 0.9944. Min Act: -0.0001
Data Index: 1257561 (C4 (Web Text))
Max Activating Token Index: 328
Click toggle to see full text
Truncated
Full Text #17
Text #18
Max Range: 1.9358. Min Range: -1.9358
Max Act: 1.0015. Min Act: -0.0001
Data Index: 951757 (C4 (Web Text))
Max Activating Token Index: 130
Click toggle to see full text
Truncated
Full Text #18
Text #19
Max Range: 1.9358. Min Range: -1.9358
Max Act: 0.9463. Min Act: -0.0001
Data Index: 991466 (C4 (Web Text))
Max Activating Token Index: 400
Click toggle to see full text
Truncated
Full Text #19