INDEX

Explanations

sections related to editing or changes made to documents

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Edit" and references to sections of content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

milo

-0.85

kamp

-0.70

ciating

-0.70

unity

-0.67

mitt

-0.66

wagen

-0.66

matically

-0.65

pher

-0.65

sonian

-0.65

bm

-0.64

POSITIVE LOGITS

 Edit

1.24

edit

0.90

âĶĢâĶĢ

0.87

Edit

0.86

 edit

0.85

 Below

0.76

.[

0.76

 Editing

0.76

 Factor

0.74

 Delete

0.72

Activations Density 0.009%

sections related to editing or changes made to documents

instances of the word "Edit" and references to sections of content

No Comments

No Known Activations