INDEX

Explanations

keywords related to publications, such as "publish," "publisher," and "pubs."

oai_token-act-pair · gpt-3.5-turbo

references to publications or media outlets

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Gaia

-0.70

 autom

-0.70

 Refuge

-0.65

OHN

-0.64

 Archangel

-0.63

FUL

-0.62

EFF

-0.62

terday

-0.61

 Trayvon

-0.61

 Reson

-0.61

POSITIVE LOGITS

lisher

1.66

lish

1.64

lishes

1.60

lishing

1.59

lique

1.54

escent

1.42

bing

1.07

lished

1.07

lik

1.07

bed

0.98

Activations Density 0.025%

keywords related to publications, such as "publish," "publisher," and "pubs."

references to publications or media outlets

No Comments

No Known Activations