INDEX

Explanations

features or characteristics associated with individuals or groups

oai_token-act-pair · gpt-3.5-turbo

words or phrases related to labels and stereotypes

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

large

-0.81

usable

-0.74

ĳ

-0.73

range

-0.73

oho

-0.71

english

-0.70

ecause

-0.69

ensive

-0.69

CLOSE

-0.68

angan

-0.68

POSITIVE LOGITS

 extraord

1.18

gery

1.05

esses

1.04

hood

1.03

doms

1.01

ry

0.95

 hordes

0.94

dom

0.92

isms

0.91

ism

0.90

Activations Density 0.350%

features or characteristics associated with individuals or groups

words or phrases related to labels and stereotypes

No Known Activations