INDEX

Explanations

locations or places within a specific region

oai_token-act-pair · gpt-3.5-turbo

instances of the letter "d."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Corvette

-0.65

 adversity

-0.62

Ley

-0.62

 Remem

-0.61

 foul

-0.59

 Laurie

-0.59

Nun

-0.59

 depreciation

-0.59

Lor

-0.56

 Ferr

-0.56

POSITIVE LOGITS

orf

1.34

ynamic

1.33

ividual

1.28

roid

1.22

aniel

1.21

yssey

1.20

ynasty

1.19

ynam

1.18

iamond

1.18

elta

1.15

Activations Density 0.056%

locations or places within a specific region

instances of the letter "d."

No Comments

No Known Activations