INDEX

Explanations

No Explanations Found

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

raviolet

-0.76

assetsadobe

-0.74

AUD

-0.73

ANG

-0.70

hyde

-0.68

beam

-0.68

 Emerald

-0.66

 Documents

-0.65

regon

-0.64

iterranean

-0.64

POSITIVE LOGITS

MpServer

0.81

¼

0.70

sty

0.66

icides

0.65

Compat

0.64

Pru

0.64

 surrog

0.63

spe

0.63

cre

0.63

 behav

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.