INDEX

Explanations

advertisements within the text

oai_token-act-pair · gpt-3.5-turbo

instances of the word "advertisement"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

hof

-0.76

ommel

-0.75

grass

-0.75

 Normandy

-0.72

rum

-0.72

lands

-0.69

oids

-0.68

peak

-0.65

fully

-0.65

gger

-0.65

POSITIVE LOGITS

 ..........

1.27

ãħĭãħĭ

1.21

ãħĭ

1.00

allery

0.95

ileaks

0.82

sembly

0.81

 teasp

0.80

veyard

0.76

aucuses

0.76

olulu

0.75

Activations Density 0.014%

advertisements within the text

instances of the word "advertisement"

No Comments

No Known Activations