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

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

Neuron 1054 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.5168. Min Act: -0.1700

Data Index: 1005947 (C4 (Web Text))

Max Activating Token Index: 624

Text #1

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.2501. Min Act: -0.1700

Data Index: 72591 (C4 (Web Text))

Max Activating Token Index: 156

Text #2

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.2029. Min Act: -0.1700

Data Index: 855258 (C4 (Web Text))

Max Activating Token Index: 434

Text #3

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.0926. Min Act: -0.1700

Data Index: 1335207 (C4 (Web Text))

Max Activating Token Index: 971

Text #4

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.1242. Min Act: -0.1700

Data Index: 48969 (C4 (Web Text))

Max Activating Token Index: 165

Text #5

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.0926. Min Act: -0.1700

Data Index: 938082 (C4 (Web Text))

Max Activating Token Index: 947

Text #6

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.0769. Min Act: -0.1700

Data Index: 378648 (C4 (Web Text))

Max Activating Token Index: 623

Text #7

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.1084. Min Act: -0.1700

Data Index: 148388 (C4 (Web Text))

Max Activating Token Index: 583

Text #8

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.0611. Min Act: -0.1700

Data Index: 1151407 (C4 (Web Text))

Max Activating Token Index: 1000

Text #9

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.0769. Min Act: -0.1700

Data Index: 1357559 (C4 (Web Text))

Max Activating Token Index: 764

Text #10

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.0769. Min Act: -0.1700

Data Index: 506848 (C4 (Web Text))

Max Activating Token Index: 361

Text #11

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.0453. Min Act: -0.1700

Data Index: 882195 (C4 (Web Text))

Max Activating Token Index: 886

Text #12

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.0295. Min Act: -0.1700

Data Index: 513311 (C4 (Web Text))

Max Activating Token Index: 647

Text #13

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.0453. Min Act: -0.1700

Data Index: 1286111 (C4 (Web Text))

Max Activating Token Index: 853

Text #14

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.0137. Min Act: -0.1700

Data Index: 1345071 (C4 (Web Text))

Max Activating Token Index: 864

Text #15

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.0453. Min Act: -0.1700

Data Index: 101355 (C4 (Web Text))

Max Activating Token Index: 626

Text #16

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.0295. Min Act: -0.1700

Data Index: 1261778 (C4 (Web Text))

Max Activating Token Index: 386

Text #17

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.0137. Min Act: -0.1700

Data Index: 1193014 (C4 (Web Text))

Max Activating Token Index: 547

Text #18

Max Range: 3.5168. Min Range: -3.5168

Max Act: 2.9821. Min Act: -0.1700

Data Index: 33808 (C4 (Web Text))

Max Activating Token Index: 847

Text #19

Max Range: 3.5168. Min Range: -3.5168

Max Act: 3.0137. Min Act: -0.1700

Data Index: 114152 (C4 (Web Text))

Max Activating Token Index: 301