INDEX

Explanations

names starting with "Hel" or related words

oai_token-act-pair · gpt-3.5-turbo

references to brands or products

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

famous

-0.69

hig

-0.68

 elig

-0.65

DragonMagazine

-0.65

 Worth

-0.63

RANT

-0.62

 ILCS

-0.60

emetery

-0.60

ppel

-0.59

hiba

-0.59

POSITIVE LOGITS

stad

0.80

agne

0.76

berger

0.69

opter

0.67

oned

0.67

iday

0.66

inki

0.64

ãĤ·ãĥ£

0.63

ãĥĺ

0.62

lain

0.62

Activations Density 0.164%

names starting with "Hel" or related words

references to brands or products

No Comments

No Known Activations