INDEX

Explanations

mentions of the location "South" and related words

oai_token-act-pair · gpt-3.5-turbo

geographical references, particularly related to locations with "outh."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ãĤ´ãĥ³

-0.78

abase

-0.77

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.74

senal

-0.70

berman

-0.69

 busted

-0.65

 Kaplan

-0.63

Mum

-0.63

ãĤ¶

-0.61

cedes

-0.61

POSITIVE LOGITS

outh

1.20

ful

1.14

ilus

0.95

ouse

0.90

some

0.87

emouth

0.87

mares

0.85

fully

0.84

ourse

0.82

fulness

0.82

Activations Density 0.011%

mentions of the location "South" and related words

geographical references, particularly related to locations with "outh."

No Comments

No Known Activations