INDEX

Explanations

keywords or phrases related to editing or publication

oai_token-act-pair · gpt-3.5-turbo

instances of editing or edits within text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Died

-0.73

 STATS

-0.70

rolet

-0.65

 Missing

-0.65

nell

-0.65

bos

-0.63

 Bicycle

-0.63

VS

-0.63

 Shank

-0.63

 Wrap

-0.62

POSITIVE LOGITS

Edited

2.31

ASC

1.67

 Edited

1.26

UTF

0.95

arious

0.78

iquette

0.78

utf

0.76

cum

0.71

ovember

0.70

ugi

0.69

Activations Density 0.005%

keywords or phrases related to editing or publication

instances of editing or edits within text

No Comments

No Known Activations