INDEX

Explanations

phrases that indicate global reach or distribution

oai_token-act-pair · gpt-3.5-turbo

references to global or international contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

rats

-0.69

rences

-0.63

arial

-0.60

ãĤ¶

-0.58

chal

-0.57

oother

-0.56

raid

-0.56

potion

-0.56

ascript

-0.56

POSITIVE LOGITS

 Europe

1.24

 America

1.20

 Asia

1.19

 globe

1.06

 Canada

1.02

 Africa

1.01

 Australia

1.00

 Britain

0.99

 Scandinav

0.97

America

0.94

Activations Density 0.112%

phrases that indicate global reach or distribution

references to global or international contexts

No Comments

No Known Activations