INDEX

Explanations

mentions of accolades or awards

oai_token-act-pair · gpt-3.5-turbo

terms and phrases related to awards, brands, and high-quality attributes

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Semitism

-0.72

 Solitaire

-0.69

 Indians

-0.65

 occupations

-0.63

 Parenthood

-0.62

 Grants

-0.62

 Females

-0.62

 Sale

-0.61

 Taxes

-0.60

 associations

-0.60

POSITIVE LOGITS

famous

1.03

uous

1.01

ishable

0.98

orious

0.98

oried

0.96

urable

0.95

acious

0.95

abled

0.95

ishly

0.93

iest

0.92

Activations Density 0.526%

mentions of accolades or awards

terms and phrases related to awards, brands, and high-quality attributes

No Comments

No Known Activations