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

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

Neuron 23 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: 7.1464. Min Range: -7.1464

Max Act: 7.1464. Min Act: -0.1700

Data Index: 321380 (C4 (Web Text))

Max Activating Token Index: 95

Text #1

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.9589. Min Act: -0.1700

Data Index: 764503 (C4 (Web Text))

Max Activating Token Index: 314

Text #2

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.6777. Min Act: -0.1700

Data Index: 700081 (C4 (Web Text))

Max Activating Token Index: 571

Text #3

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.6777. Min Act: -0.1700

Data Index: 964689 (C4 (Web Text))

Max Activating Token Index: 206

Text #4

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.5839. Min Act: -0.1700

Data Index: 339764 (C4 (Web Text))

Max Activating Token Index: 221

Text #5

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.6152. Min Act: -0.1700

Data Index: 1062841 (C4 (Web Text))

Max Activating Token Index: 54

Text #6

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.5839. Min Act: -0.1700

Data Index: 25781 (C4 (Web Text))

Max Activating Token Index: 591

Text #7

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.5527. Min Act: -0.1700

Data Index: 534421 (C4 (Web Text))

Max Activating Token Index: 896

Text #8

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.6152. Min Act: -0.1700

Data Index: 1200752 (C4 (Web Text))

Max Activating Token Index: 40

Text #9

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.6152. Min Act: -0.1700

Data Index: 815017 (C4 (Web Text))

Max Activating Token Index: 731

Text #10

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.4902. Min Act: -0.1700

Data Index: 1335757 (C4 (Web Text))

Max Activating Token Index: 177

Text #11

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.5527. Min Act: -0.1700

Data Index: 1148411 (C4 (Web Text))

Max Activating Token Index: 218

Text #12

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.6152. Min Act: -0.1700

Data Index: 270632 (C4 (Web Text))

Max Activating Token Index: 858

Text #13

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.5527. Min Act: -0.1700

Data Index: 687998 (C4 (Web Text))

Max Activating Token Index: 354

Text #14

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.5214. Min Act: -0.1700

Data Index: 1166521 (C4 (Web Text))

Max Activating Token Index: 967

Text #15

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.4589. Min Act: -0.1700

Data Index: 274803 (C4 (Web Text))

Max Activating Token Index: 402

Text #16

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.5527. Min Act: -0.1700

Data Index: 813260 (C4 (Web Text))

Max Activating Token Index: 162

Text #17

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.3964. Min Act: -0.1700

Data Index: 300111 (C4 (Web Text))

Max Activating Token Index: 918

Text #18

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.4277. Min Act: -0.1700

Data Index: 795906 (C4 (Web Text))

Max Activating Token Index: 169

Text #19

Max Range: 7.1464. Min Range: -7.1464

Max Act: 6.4902. Min Act: -0.1700

Data Index: 386294 (C4 (Web Text))

Max Activating Token Index: 212