Model: GELU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 1410 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-4l')

Text #0

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.5856. Min Act: -0.1700

Data Index: 1488690 (Python Code)

Max Activating Token Index: 151

Text #1

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.5386. Min Act: -0.1700

Data Index: 1591960 (Python Code)

Max Activating Token Index: 444

Text #2

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.5073. Min Act: -0.1700

Data Index: 513662 (C4 (Web Text))

Max Activating Token Index: 170

Text #3

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.4603. Min Act: -0.1700

Data Index: 1270014 (C4 (Web Text))

Max Activating Token Index: 650

Text #4

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.2720. Min Act: -0.1700

Data Index: 1509964 (Python Code)

Max Activating Token Index: 111

Text #5

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.1934. Min Act: -0.1700

Data Index: 71994 (C4 (Web Text))

Max Activating Token Index: 768

Text #6

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.1776. Min Act: -0.1700

Data Index: 757361 (C4 (Web Text))

Max Activating Token Index: 221

Text #7

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.1619. Min Act: -0.1700

Data Index: 1478007 (Python Code)

Max Activating Token Index: 501

Text #8

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.1304. Min Act: -0.1700

Data Index: 367430 (C4 (Web Text))

Max Activating Token Index: 578

Text #9

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.1304. Min Act: -0.1700

Data Index: 744109 (C4 (Web Text))

Max Activating Token Index: 815

Text #10

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.0831. Min Act: -0.1700

Data Index: 1050149 (C4 (Web Text))

Max Activating Token Index: 161

Text #11

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.0831. Min Act: -0.1700

Data Index: 990005 (C4 (Web Text))

Max Activating Token Index: 520

Text #12

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.0989. Min Act: -0.1700

Data Index: 817622 (C4 (Web Text))

Max Activating Token Index: 995

Text #13

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.0673. Min Act: -0.1700

Data Index: 1515153 (Python Code)

Max Activating Token Index: 174

Text #14

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.0673. Min Act: -0.1700

Data Index: 169107 (C4 (Web Text))

Max Activating Token Index: 505

Text #15

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.0515. Min Act: -0.1700

Data Index: 75139 (C4 (Web Text))

Max Activating Token Index: 1000

Text #16

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.0515. Min Act: -0.1700

Data Index: 139793 (C4 (Web Text))

Max Activating Token Index: 514

Text #17

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.0358. Min Act: -0.1700

Data Index: 1483491 (Python Code)

Max Activating Token Index: 922

Text #18

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.0358. Min Act: -0.1700

Data Index: 1318584 (C4 (Web Text))

Max Activating Token Index: 142

Text #19

Max Range: 3.5856. Min Range: -3.5856

Max Act: 3.0358. Min Act: -0.1700

Data Index: 1092200 (C4 (Web Text))

Max Activating Token Index: 919