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

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

Neuron 684 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: 5.2185. Min Range: -5.2185

Max Act: 5.2185. Min Act: -0.1698

Data Index: 75960 (C4 (Web Text))

Max Activating Token Index: 324

Text #1

Max Range: 5.2185. Min Range: -5.2185

Max Act: 5.1560. Min Act: -0.1699

Data Index: 1012531 (C4 (Web Text))

Max Activating Token Index: 27

Text #2

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.9997. Min Act: -0.1699

Data Index: 900290 (C4 (Web Text))

Max Activating Token Index: 698

Text #3

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.9997. Min Act: -0.1700

Data Index: 534017 (C4 (Web Text))

Max Activating Token Index: 78

Text #4

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.9685. Min Act: -0.1700

Data Index: 304026 (C4 (Web Text))

Max Activating Token Index: 132

Text #5

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.9685. Min Act: -0.1684

Data Index: 438079 (C4 (Web Text))

Max Activating Token Index: 255

Text #6

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.9060. Min Act: -0.1700

Data Index: 766920 (C4 (Web Text))

Max Activating Token Index: 815

Text #7

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.9060. Min Act: -0.1681

Data Index: 1350175 (C4 (Web Text))

Max Activating Token Index: 54

Text #8

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.9372. Min Act: -0.1699

Data Index: 391827 (C4 (Web Text))

Max Activating Token Index: 297

Text #9

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.9372. Min Act: -0.1700

Data Index: 1179188 (C4 (Web Text))

Max Activating Token Index: 301

Text #10

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.9060. Min Act: -0.1653

Data Index: 531124 (C4 (Web Text))

Max Activating Token Index: 728

Text #11

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.9060. Min Act: -0.1700

Data Index: 254025 (C4 (Web Text))

Max Activating Token Index: 31

Text #12

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.8435. Min Act: -0.1700

Data Index: 598160 (C4 (Web Text))

Max Activating Token Index: 320

Text #13

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.9060. Min Act: -0.1700

Data Index: 605838 (C4 (Web Text))

Max Activating Token Index: 221

Text #14

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.8435. Min Act: -0.1700

Data Index: 480270 (C4 (Web Text))

Max Activating Token Index: 124

Text #15

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.8122. Min Act: -0.1700

Data Index: 1066014 (C4 (Web Text))

Max Activating Token Index: 611

Text #16

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.8747. Min Act: -0.1685

Data Index: 48174 (C4 (Web Text))

Max Activating Token Index: 979

Text #17

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.8435. Min Act: -0.1700

Data Index: 526560 (C4 (Web Text))

Max Activating Token Index: 832

Text #18

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.8122. Min Act: -0.1700

Data Index: 311821 (C4 (Web Text))

Max Activating Token Index: 684

Text #19

Max Range: 5.2185. Min Range: -5.2185

Max Act: 4.8747. Min Act: -0.1362

Data Index: 395165 (C4 (Web Text))

Max Activating Token Index: 68