Model: GELU Model: 4 Layers, 2048 Neurons per Layer

Dataset: 80% C4 (Web Text) and 20% Python Code

Neuron 1855 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-4l')

Text #0

Max Range: 5.1060. Min Range: -5.1060

Max Act: 5.1060. Min Act: -0.1700

Data Index: 851876 (C4 (Web Text))

Max Activating Token Index: 562

Text #1

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.7935. Min Act: -0.1700

Data Index: 1262312 (C4 (Web Text))

Max Activating Token Index: 562

Text #2

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.7935. Min Act: -0.1700

Data Index: 5221 (C4 (Web Text))

Max Activating Token Index: 249

Text #3

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.7310. Min Act: -0.1700

Data Index: 698821 (C4 (Web Text))

Max Activating Token Index: 788

Text #4

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.7622. Min Act: -0.1700

Data Index: 441627 (C4 (Web Text))

Max Activating Token Index: 310

Text #5

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.7935. Min Act: -0.1700

Data Index: 929108 (C4 (Web Text))

Max Activating Token Index: 823

Text #6

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.6372. Min Act: -0.1700

Data Index: 299204 (C4 (Web Text))

Max Activating Token Index: 9

Text #7

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.6372. Min Act: -0.1700

Data Index: 1033315 (C4 (Web Text))

Max Activating Token Index: 819

Text #8

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.5747. Min Act: -0.1700

Data Index: 915544 (C4 (Web Text))

Max Activating Token Index: 290

Text #9

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.5747. Min Act: -0.1700

Data Index: 537202 (C4 (Web Text))

Max Activating Token Index: 19

Text #10

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.6372. Min Act: -0.1700

Data Index: 1206860 (C4 (Web Text))

Max Activating Token Index: 965

Text #11

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.6372. Min Act: -0.1700

Data Index: 411578 (C4 (Web Text))

Max Activating Token Index: 205

Text #12

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.5747. Min Act: -0.1700

Data Index: 39944 (C4 (Web Text))

Max Activating Token Index: 198

Text #13

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.5747. Min Act: -0.1700

Data Index: 333263 (C4 (Web Text))

Max Activating Token Index: 1001

Text #14

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.5434. Min Act: -0.1700

Data Index: 488851 (C4 (Web Text))

Max Activating Token Index: 166

Text #15

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.5434. Min Act: -0.1700

Data Index: 507968 (C4 (Web Text))

Max Activating Token Index: 737

Text #16

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.6059. Min Act: -0.1700

Data Index: 231245 (C4 (Web Text))

Max Activating Token Index: 589

Text #17

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.5434. Min Act: -0.1700

Data Index: 1004375 (C4 (Web Text))

Max Activating Token Index: 916

Text #18

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.4809. Min Act: -0.1700

Data Index: 662931 (C4 (Web Text))

Max Activating Token Index: 810

Text #19

Max Range: 5.1060. Min Range: -5.1060

Max Act: 4.5434. Min Act: -0.1700

Data Index: 13527 (C4 (Web Text))

Max Activating Token Index: 100