INDEX

Explanations

references to the Washington Post

oai_token-act-pair · gpt-3.5-turbo

references to the Washington Post

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

vana

-0.77

 Cameroon

-0.76

ework

-0.67

 assum

-0.64

âĹ¼

-0.64

pha

-0.63

cles

-0.62

hra

-0.61

cedes

-0.61

displayText

-0.61

POSITIVE LOGITS

 Post

1.42

 Examiner

1.41

 Times

1.10

 Redskins

1.10

 Capitals

1.07

post

0.98

 Monthly

0.98

Post

0.95

 Nationals

0.93

 Blade

0.93

Activations Density 0.022%

references to the Washington Post

references to the Washington Post

No Comments

No Known Activations