INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

BuyableInstoreAndOnline

-0.76

 McCabe

-0.70

ãĥ´

-0.68

 Greenberg

-0.68

ibilities

-0.68

 Judd

-0.67

Higher

-0.67

earch

-0.66

ELF

-0.64

 Â§Â§

-0.62

POSITIVE LOGITS

 toll

0.71

minist

0.70

meric

0.70

actionDate

0.67

 Tycoon

0.67

zza

0.66

 crucifix

0.62

street

0.62

star

0.61

 Toll

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.