INDEX

Explanations

No Explanations Found

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

 Dynam

-0.70

×¨

-0.68

oother

-0.68

ixture

-0.68

ystem

-0.65

ern

-0.64

 WARN

-0.63

ower

-0.62

 Defin

-0.61

 Runtime

-0.61

POSITIVE LOGITS

cox

0.75

 Cage

0.72

cius

0.64

soType

0.64

milo

0.63

 tabs

0.62

 myself

0.61

 labels

0.60

 eBay

0.60

cÃ©

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.