INDEX

Explanations

phrases indicating action or agency

oai_token-act-pair · gpt-3.5-turbo

phrases related to actions and their execution

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Flavoring

-0.73

 Founding

-0.69

ahime

-0.68

ADVERTISEMENT

-0.65

 Ratings

-0.65

FK

-0.64

]}

-0.63

 Favorite

-0.63

FontSize

-0.63

BuyableInstoreAndOnline

-0.62

POSITIVE LOGITS

 anonymously

1.22

 invis

1.06

 peacefully

1.03

 WITHOUT

1.03

 unconsciously

1.03

 collabor

1.01

 without

1.01

 indoors

0.98

 impuls

0.97

 cautiously

0.97

Activations Density 0.308%

phrases indicating action or agency

phrases related to actions and their execution

No Comments

No Known Activations