Model: GELU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 100 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-3l')

Text #0

Max Range: 5.4592. Min Range: -5.4592

Max Act: 5.4592. Min Act: -0.1700

Data Index: 408923 (C4 (Web Text))

Max Activating Token Index: 877

Text #1

Max Range: 5.4592. Min Range: -5.4592

Max Act: 5.1780. Min Act: -0.1700

Data Index: 1354609 (C4 (Web Text))

Max Activating Token Index: 317

Text #2

Max Range: 5.4592. Min Range: -5.4592

Max Act: 5.1467. Min Act: -0.1700

Data Index: 220291 (C4 (Web Text))

Max Activating Token Index: 121

Text #3

Max Range: 5.4592. Min Range: -5.4592

Max Act: 5.0842. Min Act: -0.1700

Data Index: 1279487 (C4 (Web Text))

Max Activating Token Index: 834

Text #4

Max Range: 5.4592. Min Range: -5.4592

Max Act: 5.1467. Min Act: -0.1700

Data Index: 711163 (C4 (Web Text))

Max Activating Token Index: 463

Text #5

Max Range: 5.4592. Min Range: -5.4592

Max Act: 5.1155. Min Act: -0.1699

Data Index: 553734 (C4 (Web Text))

Max Activating Token Index: 894

Text #6

Max Range: 5.4592. Min Range: -5.4592

Max Act: 5.1155. Min Act: -0.1700

Data Index: 309180 (C4 (Web Text))

Max Activating Token Index: 214

Text #7

Max Range: 5.4592. Min Range: -5.4592

Max Act: 5.1155. Min Act: -0.1700

Data Index: 732648 (C4 (Web Text))

Max Activating Token Index: 487

Text #8

Max Range: 5.4592. Min Range: -5.4592

Max Act: 5.0530. Min Act: -0.1700

Data Index: 456644 (C4 (Web Text))

Max Activating Token Index: 930

Text #9

Max Range: 5.4592. Min Range: -5.4592

Max Act: 5.0842. Min Act: -0.1700

Data Index: 646039 (C4 (Web Text))

Max Activating Token Index: 138

Text #10

Max Range: 5.4592. Min Range: -5.4592

Max Act: 5.0530. Min Act: -0.1700

Data Index: 546340 (C4 (Web Text))

Max Activating Token Index: 624

Text #11

Max Range: 5.4592. Min Range: -5.4592

Max Act: 5.0530. Min Act: -0.1699

Data Index: 1214232 (C4 (Web Text))

Max Activating Token Index: 719

Text #12

Max Range: 5.4592. Min Range: -5.4592

Max Act: 4.9905. Min Act: -0.1700

Data Index: 324034 (C4 (Web Text))

Max Activating Token Index: 975

Text #13

Max Range: 5.4592. Min Range: -5.4592

Max Act: 4.9592. Min Act: -0.1700

Data Index: 500771 (C4 (Web Text))

Max Activating Token Index: 50

Text #14

Max Range: 5.4592. Min Range: -5.4592

Max Act: 4.9280. Min Act: -0.1700

Data Index: 370205 (C4 (Web Text))

Max Activating Token Index: 852

Text #15

Max Range: 5.4592. Min Range: -5.4592

Max Act: 4.8967. Min Act: -0.1700

Data Index: 1201813 (C4 (Web Text))

Max Activating Token Index: 926

Text #16

Max Range: 5.4592. Min Range: -5.4592

Max Act: 4.9280. Min Act: -0.1700

Data Index: 169319 (C4 (Web Text))

Max Activating Token Index: 885

Text #17

Max Range: 5.4592. Min Range: -5.4592

Max Act: 4.9280. Min Act: -0.1700

Data Index: 1156663 (C4 (Web Text))

Max Activating Token Index: 148

Text #18

Max Range: 5.4592. Min Range: -5.4592

Max Act: 4.8967. Min Act: -0.1700

Data Index: 1147185 (C4 (Web Text))

Max Activating Token Index: 488

Text #19

Max Range: 5.4592. Min Range: -5.4592

Max Act: 4.8967. Min Act: -0.1700

Data Index: 964526 (C4 (Web Text))

Max Activating Token Index: 720