INDEX

Explanations

phrases related to authority figures or organizations

oai_token-act-pair · gpt-3.5-turbo

references to government agencies or affiliations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

WHERE

-0.76

dos

-0.71

ahu

-0.69

achu

-0.68

 vans

-0.67

chedel

-0.66

aza

-0.65

iphany

-0.65

anish

-0.64

soDeliveryDate

-0.64

POSITIVE LOGITS

 base

0.76

 depend

0.70

 indebted

0.70

 dearly

0.67

 depends

0.67

 sympath

0.67

 unres

0.66

Emb

0.65

suff

0.65

 sole

0.64

Activations Density 0.114%

phrases related to authority figures or organizations

references to government agencies or affiliations

No Comments

No Known Activations