INDEX

Explanations

commands or functionalities related to customization and control options within software or applications

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate functionality or capabilities of software and applications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

venant

-0.68

 Beasts

-0.68

born

-0.67

bage

-0.67

Ô

-0.66

lihood

-0.66

querque

-0.65

Øª

-0.65

ItemThumbnailImage

-0.64

CLASSIFIED

-0.63

POSITIVE LOGITS

tered

1.00

heet

0.92

tering

0.89

icia

0.81

eries

0.80

us

0.78

rack

0.77

terness

0.74

isters

0.73

sheet

0.73

Activations Density 0.018%

commands or functionalities related to customization and control options within software or applications

phrases that indicate functionality or capabilities of software and applications

No Comments

No Known Activations