INDEX

Explanations

occurrences of the word "warrant" in texts

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 Tycoon

-0.73

hedral

-0.65

ttle

-0.64

thumbnails

-0.63

este

-0.63

biz

-0.63

esta

-0.62

helle

-0.62

kens

-0.62

 Profit

-0.61

POSITIVE LOGITS

 warrant

1.14

 warrants

1.00

 authorizing

1.00

 affidavit

0.97

 authorization

0.90

 Warrant

0.89

lessly

0.83

 affidav

0.80

 exting

0.78

NING

0.77

Activations Density 0.021%

occurrences of the word "warrant" in texts

No Comments

No Known Activations