INDEX

Explanations

words related to direction or position

oai_token-act-pair · gpt-3.5-turbo

nouns and phrases related to perspectives or viewpoints

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

POST

-0.73

upper

-0.73

kn

-0.72

space

-0.70

ctrl

-0.68

click

-0.66

UP

-0.66

peer

-0.65

league

-0.64

layer

-0.64

POSITIVE LOGITS

enment

0.81

ously

0.72

 humour

0.71

DragonMagazine

0.68

 narration

0.66

iaries

0.64

 Doodle

0.64

ingly

0.64

aires

0.63

 Prelude

0.62

Activations Density 0.143%

words related to direction or position

nouns and phrases related to perspectives or viewpoints

No Comments

No Known Activations