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

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

Neuron 1844 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: 3.3899. Min Range: -3.3899

Max Act: 3.3899. Min Act: -0.1700

Data Index: 311527 (C4 (Web Text))

Max Activating Token Index: 154

Text #1

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.3585. Min Act: -0.1700

Data Index: 999148 (C4 (Web Text))

Max Activating Token Index: 995

Text #2

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.2957. Min Act: -0.1700

Data Index: 1022541 (C4 (Web Text))

Max Activating Token Index: 460

Text #3

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.2643. Min Act: -0.1700

Data Index: 798552 (C4 (Web Text))

Max Activating Token Index: 957

Text #4

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.2329. Min Act: -0.1700

Data Index: 1092784 (C4 (Web Text))

Max Activating Token Index: 321

Text #5

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.1699. Min Act: -0.1700

Data Index: 1124422 (C4 (Web Text))

Max Activating Token Index: 335

Text #6

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.1699. Min Act: -0.1700

Data Index: 302480 (C4 (Web Text))

Max Activating Token Index: 826

Text #7

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.2014. Min Act: -0.1700

Data Index: 265501 (C4 (Web Text))

Max Activating Token Index: 962

Text #8

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.2014. Min Act: -0.1700

Data Index: 884310 (C4 (Web Text))

Max Activating Token Index: 428

Text #9

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.2329. Min Act: -0.1700

Data Index: 493694 (C4 (Web Text))

Max Activating Token Index: 58

Text #10

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.1699. Min Act: -0.1700

Data Index: 47567 (C4 (Web Text))

Max Activating Token Index: 864

Text #11

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.1384. Min Act: -0.1700

Data Index: 718780 (C4 (Web Text))

Max Activating Token Index: 292

Text #12

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.1384. Min Act: -0.1700

Data Index: 84754 (C4 (Web Text))

Max Activating Token Index: 293

Text #13

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.1384. Min Act: -0.1700

Data Index: 870857 (C4 (Web Text))

Max Activating Token Index: 659

Text #14

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.1069. Min Act: -0.1700

Data Index: 463092 (C4 (Web Text))

Max Activating Token Index: 348

Text #15

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.1227. Min Act: -0.1700

Data Index: 612313 (C4 (Web Text))

Max Activating Token Index: 893

Text #16

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.1384. Min Act: -0.1700

Data Index: 986115 (C4 (Web Text))

Max Activating Token Index: 400

Text #17

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.0596. Min Act: -0.1700

Data Index: 1520185 (Python Code)

Max Activating Token Index: 181

Text #18

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.1069. Min Act: -0.1700

Data Index: 1111740 (C4 (Web Text))

Max Activating Token Index: 73

Text #19

Max Range: 3.3899. Min Range: -3.3899

Max Act: 3.1384. Min Act: -0.1700

Data Index: 117024 (C4 (Web Text))

Max Activating Token Index: 889