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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

erest

-0.71

BuyableInstoreAndOnline

-0.70

parents

-0.65

HOME

-0.65

berman

-0.65

 Bray

-0.65

Winged

-0.64

leground

-0.63

love

-0.62

 Boko

-0.61

POSITIVE LOGITS

ĸļ

0.81

obby

0.79

orship

0.78

olog

0.72

illing

0.70

uyomi

0.68

ogene

0.66

orter

0.65

oppers

0.65

alid

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.