INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

DragonMagazine

-0.68

 Nicarag

-0.65

blade

-0.64

FD

-0.63

caster

-0.62

zip

-0.61

 Apache

-0.61

 JPEG

-0.60

UCS

-0.59

BF

-0.59

POSITIVE LOGITS

uden

0.70

arten

0.69

hog

0.68

warts

0.68

ienced

0.68

iences

0.65

ichick

0.65

 tilted

0.64

nen

0.63

uana

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.