INDEX

Explanations

closures or statements indicating the end of a text segment

oai_token-act-pair · gpt-3.5-turbo

instances of the word "CLOSE."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ologically

-0.94

hid

-0.78

uana

-0.71

unal

-0.69

urg

-0.69

 alienated

-0.68

audi

-0.68

wagen

-0.68

informed

-0.67

otropic

-0.65

POSITIVE LOGITS

CLOSE

1.29

UTERS

0.90

GROUND

0.83

aign

0.80

FIX

0.79

DOWN

0.77

 Skip

0.76

INGS

0.75

 Remastered

0.73

Skip

0.73

Activations Density 0.012%

closures or statements indicating the end of a text segment

instances of the word "CLOSE."

No Comments

No Known Activations