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

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

Neuron 1995 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.4089. Min Act: -0.1700

Data Index: 44918 (C4 (Web Text))

Max Activating Token Index: 676

Text #1

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.3152. Min Act: -0.1700

Data Index: 86788 (C4 (Web Text))

Max Activating Token Index: 891

Text #2

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.2214. Min Act: -0.1700

Data Index: 1072716 (C4 (Web Text))

Max Activating Token Index: 454

Text #3

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.1902. Min Act: -0.1700

Data Index: 337968 (C4 (Web Text))

Max Activating Token Index: 803

Text #4

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.2214. Min Act: -0.1700

Data Index: 173167 (C4 (Web Text))

Max Activating Token Index: 734

Text #5

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.2214. Min Act: -0.1700

Data Index: 273349 (C4 (Web Text))

Max Activating Token Index: 435

Text #6

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.1277. Min Act: -0.1700

Data Index: 1146964 (C4 (Web Text))

Max Activating Token Index: 334

Text #7

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.1277. Min Act: -0.1700

Data Index: 585428 (C4 (Web Text))

Max Activating Token Index: 784

Text #8

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.0964. Min Act: -0.1697

Data Index: 244503 (C4 (Web Text))

Max Activating Token Index: 22

Text #9

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.0652. Min Act: -0.1700

Data Index: 464085 (C4 (Web Text))

Max Activating Token Index: 396

Text #10

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.0652. Min Act: -0.1698

Data Index: 687295 (C4 (Web Text))

Max Activating Token Index: 34

Text #11

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.0339. Min Act: -0.1700

Data Index: 508847 (C4 (Web Text))

Max Activating Token Index: 815

Text #12

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.0339. Min Act: -0.1700

Data Index: 642501 (C4 (Web Text))

Max Activating Token Index: 309

Text #13

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.0339. Min Act: -0.1700

Data Index: 201776 (C4 (Web Text))

Max Activating Token Index: 371

Text #14

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.0339. Min Act: -0.1700

Data Index: 627289 (C4 (Web Text))

Max Activating Token Index: 233

Text #15

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.0339. Min Act: -0.1700

Data Index: 1227164 (C4 (Web Text))

Max Activating Token Index: 1023

Text #16

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.0339. Min Act: -0.1690

Data Index: 1133283 (C4 (Web Text))

Max Activating Token Index: 857

Text #17

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.0027. Min Act: -0.1700

Data Index: 305433 (C4 (Web Text))

Max Activating Token Index: 328

Text #18

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.0027. Min Act: -0.1700

Data Index: 213291 (C4 (Web Text))

Max Activating Token Index: 310

Text #19

Max Range: 5.4089. Min Range: -5.4089

Max Act: 5.0027. Min Act: -0.1700

Data Index: 583019 (C4 (Web Text))

Max Activating Token Index: 655