INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

erva

-0.74

cientious

-0.71

 Claud

-0.68

 authorized

-0.68

 Byrd

-0.67

 Amendments

-0.67

 Warranty

-0.66

authorized

-0.63

anty

-0.62

onsense

-0.62

POSITIVE LOGITS

rencies

0.79

MpServer

0.78

soc

0.74

ukong

0.74

¥µ

0.72

imeters

0.70

µ

0.69

 humanities

0.67

é¾įåĸļå£«

0.67

material

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.