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

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

Neuron 1681 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: 3.6772. Min Range: -3.6772

Max Act: 3.6772. Min Act: -0.1700

Data Index: 880911 (C4 (Web Text))

Max Activating Token Index: 317

Text #1

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.5206. Min Act: -0.1700

Data Index: 1442364 (Python Code)

Max Activating Token Index: 759

Text #2

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.4893. Min Act: -0.1700

Data Index: 338092 (C4 (Web Text))

Max Activating Token Index: 288

Text #3

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.4580. Min Act: -0.1700

Data Index: 967916 (C4 (Web Text))

Max Activating Token Index: 224

Text #4

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.4736. Min Act: -0.1700

Data Index: 472787 (C4 (Web Text))

Max Activating Token Index: 775

Text #5

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.4580. Min Act: -0.1700

Data Index: 177428 (C4 (Web Text))

Max Activating Token Index: 884

Text #6

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.4423. Min Act: -0.1700

Data Index: 119519 (C4 (Web Text))

Max Activating Token Index: 581

Text #7

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.4266. Min Act: -0.1700

Data Index: 756211 (C4 (Web Text))

Max Activating Token Index: 29

Text #8

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.4109. Min Act: -0.1700

Data Index: 162148 (C4 (Web Text))

Max Activating Token Index: 669

Text #9

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.4423. Min Act: -0.1700

Data Index: 759025 (C4 (Web Text))

Max Activating Token Index: 819

Text #10

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.4109. Min Act: -0.1700

Data Index: 35285 (C4 (Web Text))

Max Activating Token Index: 340

Text #11

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.3952. Min Act: -0.1700

Data Index: 1141436 (C4 (Web Text))

Max Activating Token Index: 380

Text #12

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.3952. Min Act: -0.1700

Data Index: 1369252 (Python Code)

Max Activating Token Index: 551

Text #13

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.3795. Min Act: -0.1700

Data Index: 66062 (C4 (Web Text))

Max Activating Token Index: 913

Text #14

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.3952. Min Act: -0.1700

Data Index: 42735 (C4 (Web Text))

Max Activating Token Index: 337

Text #15

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.3952. Min Act: -0.1700

Data Index: 755182 (C4 (Web Text))

Max Activating Token Index: 926

Text #16

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.4109. Min Act: -0.1700

Data Index: 918473 (C4 (Web Text))

Max Activating Token Index: 517

Text #17

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.3952. Min Act: -0.1700

Data Index: 383417 (C4 (Web Text))

Max Activating Token Index: 601

Text #18

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.3795. Min Act: -0.1700

Data Index: 770712 (C4 (Web Text))

Max Activating Token Index: 102

Text #19

Max Range: 3.6772. Min Range: -3.6772

Max Act: 3.3795. Min Act: -0.1700

Data Index: 283880 (C4 (Web Text))

Max Activating Token Index: 539