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

 Lynd

-0.82

EStream

-0.67

 Lumpur

-0.67

 Kant

-0.63

 deduct

-0.63

NAT

-0.62

yne

-0.62

Id

-0.62

iant

-0.61

 Eston

-0.61

POSITIVE LOGITS

anni

0.83

riors

0.78

hower

0.76

aunts

0.75

isSpecialOrderable

0.75

theless

0.75

offs

0.74

ffic

0.72

oir

0.71

@@@@

0.71

Activations Density 0.000%

No Known Activations

This feature has no known activations.