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

chio

-0.83

SHIP

-0.74

yss

-0.65

catentry

-0.63

bish

-0.62

}}}

-0.61

allo

-0.61

 expressive

-0.60

omy

-0.60

 loader

-0.60

POSITIVE LOGITS

pta

0.74

Za

0.65

warts

0.63

IELD

0.63

 Blaze

0.63

 Toledo

0.62

 Plain

0.62

 Celt

0.61

cast

0.60

crow

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.