INDEX

Explanations

terms related to political entities or actions

oai_token-act-pair · gpt-3.5-turbo

instances of the placeholder token, which suggests it is looking for structural or formatting elements in the text

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 welf

-0.62

 Rover

-0.61

Wem

-0.61

 checkpoints

-0.60

 heights

-0.60

wip

-0.59

 streak

-0.57

 beginnings

-0.57

gypt

-0.57

 Emerson

-0.56

POSITIVE LOGITS

venient

1.42

secut

1.39

cerned

1.34

stant

1.33

cern

1.31

crete

1.30

ventional

1.29

cept

1.28

verted

1.27

structed

1.27

Activations Density 0.033%

terms related to political entities or actions

instances of the placeholder token, which suggests it is looking for structural or formatting elements in the text

No Comments

No Known Activations