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

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

Neuron 1692 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: 6.6236. Min Range: -6.6236

Max Act: 6.6236. Min Act: -0.1700

Data Index: 1092559 (C4 (Web Text))

Max Activating Token Index: 802

Text #1

Max Range: 6.6236. Min Range: -6.6236

Max Act: 6.6236. Min Act: -0.1700

Data Index: 869124 (C4 (Web Text))

Max Activating Token Index: 1021

Text #2

Max Range: 6.6236. Min Range: -6.6236

Max Act: 6.5299. Min Act: -0.1700

Data Index: 506880 (C4 (Web Text))

Max Activating Token Index: 923

Text #3

Max Range: 6.6236. Min Range: -6.6236

Max Act: 6.4986. Min Act: -0.1700

Data Index: 694896 (C4 (Web Text))

Max Activating Token Index: 897

Text #4

Max Range: 6.6236. Min Range: -6.6236

Max Act: 6.3111. Min Act: -0.1700

Data Index: 1325532 (C4 (Web Text))

Max Activating Token Index: 103

Text #5

Max Range: 6.6236. Min Range: -6.6236

Max Act: 6.1080. Min Act: -0.1700

Data Index: 447361 (C4 (Web Text))

Max Activating Token Index: 192

Text #6

Max Range: 6.6236. Min Range: -6.6236

Max Act: 6.1392. Min Act: -0.1700

Data Index: 1359333 (C4 (Web Text))

Max Activating Token Index: 133

Text #7

Max Range: 6.6236. Min Range: -6.6236

Max Act: 6.2330. Min Act: -0.1700

Data Index: 241300 (C4 (Web Text))

Max Activating Token Index: 683

Text #8

Max Range: 6.6236. Min Range: -6.6236

Max Act: 6.1236. Min Act: -0.1700

Data Index: 1223731 (C4 (Web Text))

Max Activating Token Index: 615

Text #9

Max Range: 6.6236. Min Range: -6.6236

Max Act: 6.0924. Min Act: -0.1700

Data Index: 633905 (C4 (Web Text))

Max Activating Token Index: 231

Text #10

Max Range: 6.6236. Min Range: -6.6236

Max Act: 6.0299. Min Act: -0.1699

Data Index: 1248629 (C4 (Web Text))

Max Activating Token Index: 818

Text #11

Max Range: 6.6236. Min Range: -6.6236

Max Act: 5.9986. Min Act: -0.1699

Data Index: 1458041 (Python Code)

Max Activating Token Index: 129

Text #12

Max Range: 6.6236. Min Range: -6.6236

Max Act: 5.9049. Min Act: -0.1700

Data Index: 384254 (C4 (Web Text))

Max Activating Token Index: 113

Text #13

Max Range: 6.6236. Min Range: -6.6236

Max Act: 5.9517. Min Act: -0.1700

Data Index: 507515 (C4 (Web Text))

Max Activating Token Index: 132

Text #14

Max Range: 6.6236. Min Range: -6.6236

Max Act: 5.9674. Min Act: -0.1699

Data Index: 469863 (C4 (Web Text))

Max Activating Token Index: 450

Text #15

Max Range: 6.6236. Min Range: -6.6236

Max Act: 5.9830. Min Act: -0.1700

Data Index: 337404 (C4 (Web Text))

Max Activating Token Index: 711

Text #16

Max Range: 6.6236. Min Range: -6.6236

Max Act: 5.9517. Min Act: -0.1700

Data Index: 316192 (C4 (Web Text))

Max Activating Token Index: 615

Text #17

Max Range: 6.6236. Min Range: -6.6236

Max Act: 5.9361. Min Act: -0.1700

Data Index: 262863 (C4 (Web Text))

Max Activating Token Index: 779

Text #18

Max Range: 6.6236. Min Range: -6.6236

Max Act: 5.8267. Min Act: -0.1700

Data Index: 970411 (C4 (Web Text))

Max Activating Token Index: 921

Text #19

Max Range: 6.6236. Min Range: -6.6236

Max Act: 5.8580. Min Act: -0.1700

Data Index: 835934 (C4 (Web Text))

Max Activating Token Index: 66