INDEX

Explanations

words related to public opinions and political stances

oai_token-act-pair · gpt-3.5-turbo

phrases and statistics related to public opinion on political or social issues

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

ĸļ

-0.89

©¶æ¥µ

-0.79

»Ĵ

-0.70

 Thieves

-0.70

 Miracle

-0.70

Luck

-0.68

 Complex

-0.68

Capt

-0.68

 Maze

-0.67

ĨĴ

-0.65

POSITIVE LOGITS

 support

1.60

 opposes

1.60

 oppose

1.56

 opposed

1.49

 supports

1.49

 supported

1.45

 SUPPORT

1.44

support

1.42

 supporting

1.42

 backing

1.40

Activations Density 0.679%

words related to public opinions and political stances

phrases and statistics related to public opinion on political or social issues

No Comments

No Known Activations