INDEX

Explanations

phrases related to opinions or perspectives

oai_token-act-pair · gpt-3.5-turbo

references to collective opinions or perceptions

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

Not in Any Lists

No Comments

Negative Logits

arette

-0.71

ibal

-0.71

Address

-0.68

odore

-0.68

Anything

-0.67

ococ

-0.66

 Alone

-0.66

yrinth

-0.66

 Anything

-0.65

illus

-0.65

POSITIVE LOGITS

 observers

1.41

 commentators

1.30

 pundits

1.25

 critics

1.19

 analysts

1.19

 onlook

1.19

 commenters

1.15

 detractors

1.05

 viewers

1.05

 readers

1.03

Activations Density 0.161%

phrases related to opinions or perspectives

references to collective opinions or perceptions

No Comments

No Known Activations