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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ibly

-0.80

 Illum

-0.71

antry

-0.68

aped

-0.68

aciously

-0.66

imedia

-0.64

rooms

-0.64

ingly

-0.63

APE

-0.63

Native

-0.63

POSITIVE LOGITS

kowski

0.72

sein

0.69

£ı

0.68

zynski

0.67

 Gerr

0.67

 indict

0.66

Ri

0.66

hower

0.64

 persecut

0.62

 disse

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.